Multilingual human-computer interaction method and system based on numerical‑sequence final interface

Abstract

The present invention relates to the technical field of human-computer interaction, and specifically relates to a multilingual human-computer interaction method and system based on a numerical‑sequence final interface. By using numerical‑sequence finals as a core adaptation foundation, a parent-child interface as a collaborative core, and an extended interface as a functional carrier, character layout is bound to the numerical‑sequence finals, thereby constructing a unified human-computer interaction system integrating use, learning, and teaching across Chinese, Chinese-English bilingual, or multilingual contexts. The present invention is applicable to various types of electronic devices such as mobile terminals (smart phones, tablet computers), smart wearable devices (smart watches, bracelets), commercial devices (POS machines, cash register terminals), industrial control devices (industrial control panels, operation consoles), and office devices (desktop computers, notebook computers), and can cover diverse scenarios such as daily communication input, professional text creation, language teaching, cross-border business communication, industrial instruction input, and children's enlightenment education, and is particularly suitable for complex human-computer interaction requirements such as multilingual tonal input, cross-device seamless switching, and multi-modal interaction.

Description

A multilingual human-computer interaction method and system based on number sequence vowel interface Technical Field

[0001] This application relates to the field of human-computer interaction technology, and more specifically, to a multilingual human-computer interaction method and system based on a number sequence vowel interface. Background Technology

[0002] 1. Application No. CN202510037752.3 "Human-computer interaction method with hardware and software keypad and / or extended interface";

[0003] 2. The "Hard and soft keyboard and / or extended interface integrated human-computer interaction method for teaching and learning" in application number CN202510245528.3 is the source of number sequence vowels, tones and character layout and the basis for sharing number sequence candidate sequences among them;

[0004] 3. Patent No. CN110865717B, “Intelligent Chinese and English Input System for Embedded Platforms”.

[0005] This application represents an improvement and refinement of Background Technology 1, an extension and expansion of Background Technology 2 on a digital keyboard or interface within the same human-computer interaction system, or an improvement and refinement of intelligent Chinese and English input systems such as Background Technology 3. The previously unpublished application, combining Background Technology 1 and 2, has constructed a unified human-computer interaction technology system for Chinese, Chinese-English bilingual, and multilingual systems—large and / or small keyboards and their extended interfaces, whether software, hardware, or a combination of both, can not only work together (provided the numerical vowels are the same, when inputting the same initial consonant / initial / English letter, the candidate string is consistent), but also interchangeable (if the numerical vowels are the same and the numerical string layout and order are consistent, a mother-child interface is formed, only the input method for the main / consonant set differs). The mother interface has more layout options (as shown in Figures 1-4 and Figures 44 or 42 respectively), and has a many-to-one subordinate relationship with the child interfaces. However, the child interfaces are compatible with more portable devices, and users often use the child interfaces first, switching to the mother interface when more interaction is needed, resulting in a perceptual reverse correspondence.

[0006] With the acceleration of globalization and the increasing diversification of electronic devices, the demand for human-computer interaction has evolved from single-language input to a comprehensive system that adapts to multiple devices, is compatible with multiple languages, covers multiple scenarios, and is multimodal collaborative. However, existing input method technologies face multiple structural defects in practical applications. At the device adaptation level, there are significant gaps in keyboard layout and input logic across different terminal devices such as desktops, mobile terminals, and smart wearables. Full-size keyboards have low character utilization on large-screen devices, while small-screen devices suffer from high accidental touch rates due to crowded keys. Users must repeatedly adapt to operating habits when switching between devices, severely compromising input continuity. Furthermore, specialized equipment such as industrial control systems lacks targeted adaptation solutions, making it difficult to effectively implement core functions. Regarding tone annotation technology, existing solutions have significant functional limitations. Numerical tone and character tone annotation methods are isolated, failing to achieve flexible switching between toneless and tone-based input, and struggling to simultaneously meet the dual needs of rapid daily input and precise input in professional scenarios. The lack of a unified mapping logic for multilingual tone rules leads to inefficient and frequently conflicting tone-based language input. In terms of interaction, existing technologies rely excessively on keyboard input, exhibiting weak adaptability to scenarios with busy hands, special populations, and non-digital information. Although some solutions incorporate voice or image recognition functions, multimodal interaction response delays are significant, logical disconnects are severe, and recognition accuracy drops significantly in complex environments, failing to form an organically collaborative system. The design of extended functions suffers from fundamental contradictions; core input functions are forcibly bound to scenario-based extensions, narrowing the scope of patent protection and making it easier to circumvent, while also creating interface redundancy and reducing input efficiency. In the area of ​​multilingual support, each language's encoding system is independent, leading to prominent issues of character key overlap and logical conflicts during bilingual input. There is a lack of a common adaptation framework between toned and non-tonal languages, limiting the efficiency of cross-border communication and multilingual document creation. Furthermore, existing technologies lack a parent-child interface collaboration mechanism; different keyboard types cannot achieve candidate sequence consistency and interface interchangeability, resulting in low cross-device collaboration efficiency. Multiple input modes, such as abbreviated pinyin, full pinyin, and double pinyin, operate in isolation, lacking unified core logic support. Mode switching requires manual operation and cannot be mutually converted, leaving the bottleneck of input efficiency for initial consonants unresolved for a long time. These shortcomings collectively lead to fragmented input experiences, high learning costs, and insufficient scenario coverage, severely restricting the practicality and universality of human-computer interaction technology. The root of all these problems lies in the fact that existing technologies are all based on the character layout of the T9 keyboard, arranged only according to the alphabetical order of English letters, never taking into account the combination and pairing rules of various phonemes such as initials, medials, finals, and tones in Chinese Pinyin. This results in an unreasonable layout of phonemes and their input order, with systemic original design flaws that no longer meet the needs of the times.Specifically, existing technologies cannot build a cross-device compatible digital key system, making it difficult to achieve single-key triggering of vowels and a minimalist layout for small-screen devices; the lack of a unified mapping table allows key configuration and interface construction to be arbitrarily modified, disrupting the consistency of input logic; the absence of a parent-child interface system causes cross-device input interruptions and fails to maintain contextual continuity; fragmented multi-language encoding logic leads to frequent conflicts during bilingual mixed input, and it is impossible to adapt to languages ​​with tones through a unified framework; core technical features are easily split and circumvented, and existing solutions cannot effectively identify and intercept equivalent input behaviors that bypass the core mapping table or collaborative logic.

[0007] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0008] The purpose of this application is to provide a multilingual human-computer interaction method and system based on a number sequence vowel interface, which can achieve cross-device input logic consistency, support multilingual mixed input without conflict, improve input efficiency, and effectively prevent technical circumvention.

[0009] The overall purpose of this application is to integrate the use, learning, and teaching of English and phonetic characters and sentences using Chinese, Chinese-English bilingual, or Chinese-English and multilingual hardware and / or static / dynamic interfaces, extended interfaces, or multimodal graphical user interfaces, or to replace them with mutually cooperating hardware and / or static / dynamic interfaces, extended interfaces, or multimodal graphical user interfaces, to input and prompt consistent candidate pinyin and / or example characters, character strings, and / or sound strings. This aims to systematically construct a human-computer interaction interface, human-computer interaction method, and conventional or multimodal human-computer interaction information processing system that integrates the use, learning, and teaching of Chinese phonetic characters and sentences and Chinese-English bilingual English phonetic characters and sentences, both domestically and internationally, online and offline. Furthermore, it prepares for the integration of use, learning, and teaching of human-computer interaction interfaces in Chinese-English and multilingual languages.

[0010] The key technology of this application is the character layout of a digital keyboard for English and Pinyin symbols and / or example characters, which is compatible with the character layout of an alphabetic keyboard to form a parent-child keyboard system. English characters are isotropic and can be combined arbitrarily; however, the phonemes of Pinyin, such as initials, medials, finals, and tones, can only be combined selectively. Existing technologies are mainly based on English keyboards, while this application is mainly based on Pinyin keyboards, thus achieving the best of both worlds and complementing each other: first, a completely new Chinese or Chinese-English bilingual hard keyboard, soft keyboard, or a combination of hard and soft digital keyboards (collectively referred to as hard and soft keypads) are designed according to the initial consonant order or alphabetical order; then, it is further extended, expanded, or replaced with static and dynamic interfaces, extended interfaces, or multimodal graphical user interfaces, so that small touchscreens such as mobile phones can also be used for children's Pinyin and literacy enlightenment in Chinese or Chinese-English bilingual teaching.

[0011] This innovative solution provides a comprehensive suite of human-computer interaction information processing capabilities for both learning and teaching, encompassing input, retrieval, and real-time online and offline interactive teaching in both Chinese and English using hardware / soft keyboards and / or static / dynamic interfaces, extended interfaces, or multimodal graphical user interfaces. It can handle English and phonetic / word-based human-computer interaction information processing using hardware / soft keyboards and / or static / dynamic interfaces, extended interfaces, or multimodal graphical user interfaces, or further supplement with technologies such as speech recognition input and speech synthesis output for conventional human-computer interaction information processing. Alternatively, it can utilize local or remote generative artificial intelligence large language models to support text and speech processing using hardware / soft keyboards and / or static / dynamic interfaces, extended interfaces, or multimodal graphical user interfaces, or further combine static or dynamic graphics, or further combine audio, video, and multimodal information such as time and space, to handle Chinese, Chinese-English bilingual, or multilingual human-computer interaction information processing, or combine these technologies.

[0012] The technical solution of this application is as follows:

[0013] 1. A multilingual human-computer interaction method based on a number-sequence vowel interface, characterized in that it includes:

[0014] 1) Define a unique set of numerically ordered finals and mapping rules: The numerically ordered finals are formed by splitting or merging the medial finals of Chinese Pinyin according to the principle of "consistency of tongue position in pronunciation", with a total of 11 core finals (ɑ, ɑnɡ, engɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi). Each set of finals corresponds to a unique digital key position, and a unique mapping table of "final - key position identifier - encoding sequence" is established. The mapping table cannot be modified by conventional software configuration.

[0015] 2) Configure a cross-device compatible numeric keypad and extended keypad system: Extended keypad identifiers are selected from "11" and "12", and extended characters are selected from one or more of "-", "=", "*", and "#". Among them, "11" and "12" are specifically used to adapt to the keypad extension of small-screen smart wearable devices such as smartwatches. The extended keypad identifiers and core numeric keypads (1-0) are included in the "vowel-keypad identifier-encoding sequence" three-in-one mapping table, and their mapping relationship cannot be modified or split separately. Different devices (mobile terminals, smart wearables, industrial control equipment) share the same numeric vowel mapping rules and candidate prompt sequences, and achieve seamless synchronization of mapping rules through USB HID protocol and Bluetooth HID protocol, with a synchronization delay of ≤50ms.

[0016] 3) Construct an irreplaceable numerical vowel interface: The interface contains 10-16 keys (including "11" and "12" extended keys). The key layout follows the "high-frequency vowels close to the hand principle" and the "minimalist layout principle for small screen devices". Small screen devices such as smartwatches prioritize the use of "11" and "12" extended keys to expand the core vowels and achieve single-key triggering of vowels. The interface layout is strongly bound to the numerical vowel mapping rules, and the correspondence between the core keys and vowels cannot be modified through the conventional interface customization function. The key layout of small screen devices such as smartwatches needs to be synchronized with the parent-child interface system to ensure the consistency of input logic across devices.

[0017] 4) Construct a collaborative parent-child interface system: The keyboard and its extended interface with the same number sequence vowels and consistent number sequence string layout and sorting constitute the parent-child interface. They share the same encoding mapping table and candidate word library. When the same initial consonant combination is input, the same candidate list is output. The control module realizes the zero-delay switching of the parent / child interface (switching delay ≤30ms) and keeps the input context uninterrupted during the switching process.

[0018] 5) Multilingual unified encoding logic: Based on the core encoding rules of the numerical vowels, unified encoding of Chinese, English and other languages ​​with tones is achieved through "vowel reserved encoding position + speech feature mapping". Bilingual mixed input can be completed without switching language modes, and there is no conflict between the encoding sequences of different languages.

[0019] 6) Core Technology Differentiation and Indivisible Features: The method does not rely on the character mapping relationship of the existing English keyboard, but achieves input through an independent numerical vowel encoding system, which is different from the existing input method technology based on letter key reuse; and the aforementioned technical features 1)-5) are an indivisible whole for realizing the core function of this method, which needs to be implemented in a coordinated manner. Any scheme that breaks it down into independent steps or modules and only implements some features falls within the protection scope of this method; furthermore, the identification standard and technical logic for the splitting of core technical features are as follows: if a scheme breaks down "vowel mapping - key configuration - interface construction - collaborative interaction - encoding" into an indivisible whole, it will be considered as follows: Any attempt to bypass the core mapping table or collaborative logic to achieve equivalent input functionality by modifying the core mapping relationship or splitting the association between extended keys and core keys under the guise of "device-specific customization" is considered a splitting circumvention. The technical identification of splitting circumvention is achieved by detecting the temporal continuity of input operations (core operation interval ≤ 100ms) and the completeness of the encoding sequence (whether any link in the three-in-one mapping table is missing). The number sequence vowel encoding system of this method is a unified core framework; all cross-device adaptations, scenario expansions, and mode optimizations must be implemented based on this framework, and independent input logic cannot be built outside of it.

[0020] 2. The method as described in technical solution 1, characterized in that the specific configuration of the number sequence vowel interface includes any of the following types:

[0021] 1) 11-key basic vowel interface: It consists of the keys corresponding to 1-9, 0 (corresponding to 10) and the extended character "-" (marked as 11);

[0022] 2) 10-key compressed vowel interface: Combines the vowel combinations corresponding to two adjacent keys in the 11-key interface, consisting of keys 1-9 and 0;

[0023] 3) 12-key multilingual vowel interface: Based on the 11-key interface, a new key (marked as 12) corresponding to the extended character "=" is added, reserving space for multilingual encoding;

[0024] 4) 15-key multilingual vowel interface: Add 4 character tone keys T1-T4 to the 11-key interface, or use a shared key layout for "punctuation-tone";

[0025] 5) 16-key full-function character layout: Add 4 tone keys T1-T4 to the 12-key layout, or use a shared key layout for "punctuation-tone".

[0026] 3. The method described in technical solution 1, characterized in that the specific construction method of the set of numerically ordered finals is as follows: the medial finals other than uo and üe are split into medial finals + finals, and merged into 11 sets of adapting finals (ɑ, ɑnɡ, engɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi), wherein the positions of ɑ and u / o / uo / i are fixed, en is only interchanged with e / üe / er, and the remaining finals can be interchanged arbitrarily; when there is no need for adaptation sharing, the position of each final is not restricted; two sets of finals can be merged into 10 sets of compressed numerically ordered finals.

[0027] 4. The method described in technical solution 1, characterized in that the sorting of the 11 groups of numerical vowels can be selected from any of the following schemes: (1) Efficient input scheme: high-frequency vowels (ɑ, ɑn, en) are bound to keys 1-3, and low-frequency vowels (ei / ü, ou, ɑi) are bound to keys 9, 0, 11; (2) Letter keyboard compatible scheme: sorted according to the order of the letter keyboard digital keys; (3) Children's teaching scheme: sorted according to pronunciation difficulty (simple vowels → compound vowels → nasal vowels).

[0028] 5. The method described in technical solution 1, characterized in that the dynamic keyboard supports layered display: the basic layer (small screen ≤ 3.5 inches) retains only 1-12 keys (including extended keys 11 and 12), occupying ≤ 30% of the screen height; the standard layer (medium screen 3.5-7 inches) includes all keys + tone keys + mode switching keys, occupying 40%-50% of the screen height; the extended layer (large screen ≥ 7 inches) supports shortcut key binding + reserved identifiers + single-key dual-phoneme candidate area, occupying 50%-60% of the screen height.

[0029] 6. The method described in technical solution 1, characterized in that the dynamic keyboard and the external hard keyboard are automatically detected via USB HID protocol and Bluetooth HID protocol, with a detection delay of ≤100ms; it supports automatic hiding / synchronous display / floating mode switching, with a switching delay of ≤50ms; when the hard keyboard inputs a single key dual-phoneme combination, the corresponding key on the dynamic keyboard is synchronously highlighted; for small-screen devices such as smartwatches, touch optimization is added for the "11" and "12" extended keys: it supports adaptive touch area (minimum touch area ≥8×8mm), accidental touch filtering (the pressure threshold can be customized, ranging from 5-20g), and touch recognition accuracy ≥98% (test environment: room temperature 25℃±5℃, humidity 40%-60%, small-screen device screen size ≤2 inches, processor ≥Snapdragon 855 / Apple A13, memory ≥2GB).

[0030] 7. A method for extending the tone of a number sequence based on the method described in technical solution 1, characterized in that it includes:

[0031] 1) Single tone function: After inputting the initial-medial-final combination, the numbers 1-4 correspond to the first-fourth tones of Chinese, and the number 5 corresponds to the neutral tone or no tone. The candidate list of syllables with tone is displayed simultaneously.

[0032] 2) Number sequence dual tone function: Add a set of number sequence tone configurations. The number 6 corresponds to the neutral tone or no tone, and the numbers 7-0 correspond to the fourth tone and the first to third tone of Mandarin, respectively. It supports custom sorting and forms a dual tone system with the number sequence single tone. It can simultaneously prompt two sets of candidate syllables with tone.

[0033] 3) Candidate prompt rules: The extended interface determines the maximum number of prompt strings per page based on the number sequence vowel, key position identifier and extended key position identifier, number sequence string, multilingual characters and screen size.

[0034] 8. The method described in technical solution 7 is characterized in that the custom sorting of the numerical dual tones supports the user setting of a "high-frequency tone priority" rule: in adult mode, the four tones of Chinese (usage rate ≥35%) are bound to the 7 key, and in child mode, the first tone and the neutral tone (total usage rate ≥40%) are bound to the 6-7 key. The custom rule can be synchronized to the parent-child interface, and the tone input error rate is ≤3%.

[0035] 9. The method described in technical solution 7, characterized in that the number of string prompts on each page of the extended interface is dynamically adjusted according to the screen size: (1) Small screen devices (screen size ≤ 2 inches, including smartwatches): 2-3, character size ≥ 16, supporting up and down swiping to switch pages; (2) Small screen devices (screen size 2-3.5 inches): 3-5, character size ≥ 14; (3) Medium screen devices (3.5-7 inches): 6-8, character size ≥ 12; (4) Large screen devices (≥ 7 inches): 9-11, supporting left and right swiping to switch pages; the candidate prompt rules for all devices are strongly bound to the number sequence vowel mapping rules and cannot be modified independently.

[0036] 10. A dual-mode tone interface human-computer interaction method based on the method described in technical solution 1, characterized in that it includes:

[0037] 1) Character tone configuration: Includes independent tone keys and shared keys. The size of the independent keys is the same as that of the numeric keys, which can be adapted to a 4-row 4-column or circular layout; the shared key mapping relationship is ",-T1, .-T2, ?-T3, !-T4" and / or ",-T1, .-T2, / -T3, ]-T4 or \-T4";

[0038] 2) Seamless expansion of consonant set: No need to switch input modes, supports two input rules - ① initial consonant + medial vowel + character tone + final vowel; ② initial consonant + character tone + medial vowel + final vowel, sharing the number sequence final vowel mapping and candidate logic with the main consonant set;

[0039] 3) Dual-mode collaboration: Number tone and character tone can be switched freely, and a tone combination vocabulary mapping library is shared. When the same tone identifier is input, the candidate results are consistent. When there is a conflict, priority is given to the identification and the operation is triggered first. The system automatically remembers the most recently used tone mode.

[0040] 11. The method as described in technical solution 10, characterized in that the triggering logic of the shared key for "punctuation-tone" is as follows: a light touch (press duration ≤ 200ms) outputs a punctuation mark, a long press (press duration ≥ 300ms, child mode ≥ 400ms) pops up a tone candidate list (T1-T4), and the tone is output by sliding to the corresponding tone key and releasing it during the long press; the user can customize the duration threshold of the light touch / long press (range 100-500ms).

[0041] 12. The method as described in technical solution 10, characterized in that the consonant set expansion is automatically activated in the following scenarios:

[0042] (1) If no further input is received within 500ms after inputting “initial consonant + medial consonant”, the consonant set candidate will automatically pop up;

[0043] (2) In multilingual mixed input scenarios, automatic parsing is performed according to consonant set rules;

[0044] (3) In children's teaching scenarios, the consonant set candidates are displayed in a synchronous animation of the pinyin splitting of "initial consonant + medial vowel + tone".

[0045] 13. A generalized dynamic extension interface human-computer interaction method based on the method described in technical solution 1, characterized in that it includes:

[0046] 1) Flexible wake-up methods: Supports shortcut keys, voice commands, gestures, floating icons, or user-defined trigger methods. On Android 8.0 and iOS 12.0 and above mobile terminals, the trigger response latency is ≤80ms (test environment: room temperature 25℃, no external interference); the voice wake-up word can be customized, and the recognition accuracy is ≥92% in ambient noise below 50dB and in the frequency range of 200-3000Hz. Offline wake-up is supported.

[0047] 2) Core interface configuration: Includes associated prompt area, dynamic candidate area, fault-tolerant prompt area, and multimodal interaction area. The multimodal interaction area is compatible with the input logic mapping of hard keyboard, static soft keyboard, and dynamic soft keyboard.

[0048] 3) Multimodal Interaction Extension: Supports multimodal input and output such as speech-to-text, text-to-speech, gesture control, and image character recognition. Multimodal data needs to be mapped to the corresponding numerical vowel codes, which are then used in conjunction with key input codes to generate candidate data and synchronized to the dynamic candidate area in real time. Among them, speech-to-text supports ≥8 commonly used languages, with an online recognition accuracy of ≥95% and an offline recognition accuracy of ≥90%. Image character recognition supports multilingual characters with an accuracy of ≥90%.

[0049] 4) Adaptation and optimization: The interface can be dragged and scaled (range 50%-200%), the font size of the dynamic candidate area is adjustable from 8 to 24 points, and it supports switching between dark / light / high contrast modes; the dynamic candidate area and the trigger operation area are horizontally aligned, with an alignment error of ≤1 pixel.

[0050] 14. The method described in technical solution 13, wherein the fault-tolerant prompt area calculates the probability of mis-touch / mis-recognition based on the physical distance of the key position or the similarity of voice recognition, the accuracy of the Top1 correction candidate is ≥92%, and the correction logic is consistent with the number sequence vowel mapping.

[0051] 15. The method described in technical solution 13 is characterized in that the multimodal interaction area mode adaptation is as follows: (1) Adult mode: enhanced professional terminology recognition, accuracy ≥ 90%; (2) Child mode: enhanced pronunciation error tolerance, non-standard pronunciation recognition accuracy ≥ 88%, picture book character recognition accuracy ≥ 90%; (3) Single key dual phoneme collaboration: after speech recognition with syllable adjustment, the corresponding key position of the dynamic keyboard is synchronously highlighted.

[0052] 16. A method for scene adaptation, multi-directional layout adaptation, and core engine-supported human-computer interaction based on the method described in technical solution 1, characterized in that it includes:

[0053] 1) Cross-device adaptation: Configure the effective touch area of ​​the key surface according to the device type. Mobile terminals ≥ 8×8mm, smart wearable devices (including smartwatches) ≥ 10×10mm ("11" and "12" extended keys are the same as the standard), commercial devices ≥ 12×12mm, industrial control panels ≥ 15×15mm; the core input logic and number sequence vowel mapping rules of all devices remain completely consistent. Only the physical size and layout density of the keys can be adjusted. The core coding association relationship cannot be modified.

[0054] 2) Multi-directional layout adaptation: Supports right-angle left vertical, right-angle right vertical, and top reverse arrangement. The character direction of the top reverse arrangement can be selected by the user. It supports manual switching or automatic adaptation according to the document layout direction, and the switching delay is ≤50ms.

[0055] 3) Scene customization and adaptation: Children's teaching scenarios are equipped with pronunciation and mouth shape animation display, 3-level speech speed adjustment and follow-up reading feedback function; industrial control scenarios optimize core command input and caching; cross-border communication scenarios enhance multilingual offline translation and pronunciation prompts based on number sequence vowel encoding;

[0056] 4) Multilingual extension and adaptation: Based on speech features, core characters of other languages ​​are bound to reserved encoding positions of number vowels, supporting standard pronunciation prompts for words in multiple languages ​​(naturalness ≥ 4.0 points);

[0057] 5) Core Support Engine: Integrates offline encoding verification based on number-sequential vowel encoding, mixed-mode parsing, offline dictionary indexing, and multilingual bidirectional offline translation functions. The offline dictionary contains ≥80,000 entries for Chinese, ≥50,000 entries for English, and ≥30,000 entries / language for other languages.

[0058] 17. The method described in technical solution 16, characterized in that the visual teaching function of the children's teaching scenario includes: animated display of the pronunciation mouth shape of Chinese Pinyin, 3-level speech speed adjustment (50 / 100 / 150 words / minute), syllable follow-up interactive game, giving encouraging feedback after correct input, and single-key dual-phoneme combination supporting visual split display.

[0059] 18. The method described in technical solution 16, characterized in that the offline encoding verification module of the core support engine rejects invalid combinations such as "zh+ü" and "b+iong", the verification response delay is ≤150ms (test environment: room temperature 25℃, no external interference), and when it fails, it prompts "recommended initial and medial combination list" in a fixed order.

[0060] 19. The method described in technical solution 16 is characterized in that, in cross-border communication scenarios, it supports bidirectional offline translation between Chinese and English and languages ​​with tones; the translation accuracy of commonly used phrases is ≥92%, the translation accuracy of professional terms is ≥88%, and the translation response time is ≤300ms; single-key dual-phoneme combination rules are reused to languages ​​with tones, and the bilingual mixed tone input conflict rate is ≤3%.

[0061] 20. The method described in technical solution 16 is characterized in that the core instructions for industrial control scenarios support offline caching (cache quantity ≥ 1000, retention time ≥ 7 days), and the input time of core instructions is ≤ 1 second; after integrating the single-key dual-phoneme combination, the input time of the instruction with tone is ≤ 180ms / instruction, the offline cached combination instructions are ≥ 800 groups, and the availability after network disconnection is ≥ 99.8%.

[0062] 21. A single-key dual-phoneme enhanced human-computer interaction method based on the method described in technical solution 1, characterized in that it includes:

[0063] 1) Single-key double-phoneme combination type: ① Medial vowel and number single tone / character tone combination (i + 1-4 tone / neutral tone, u / ü + 1-4 tone / neutral tone); ② Tone and medial vowel combination (1-4 tone / neutral tone + i, 1-4 tone / neutral tone + u / ü, tone order can be adjusted);

[0064] 2) Single-key triggering logic and indivisibility: All dual-phoneme combinations are triggered by a single key (0 key, bound to the medial i) / "-" key (bound to the medial u / ü). The triggering method includes two irreplaceable implementation paths: ① Long press of a digital key (long press duration ≥ 300ms, customizable threshold range 100-500ms); ② Shortcut key combination triggering: Preset combination keys are locked through system permission requests to prevent third-party software from occupying them. Triggering occurs directly through the system's underlying interface, without being intercepted by third-party software. Dual-key input is triggered by a short press (short press duration...). The system distinguishes the triggering method by dual judgment of press duration and signal characteristics (≤200ms), with no conflict and a trigger response delay of ≤50ms; and the encoding sequence of single-key dual-phoneme combination forms a unique correspondence with the core encoding system of number sequence vowels, which cannot be replaced by equivalent means; the aforementioned "single-key triggering-encoding mapping-cross-mode compatibility" constitutes an indivisible functional whole, and it is prohibited to circumvent it by splitting "triggering operation and encoding generation" into independent steps, that is, any scheme that first triggers the key and then generates an equivalent dual-phoneme code through an additional software module falls within the protection scope of this method;

[0065] 3) Cross-mode compatibility: The single-key dual-phoneme combination rule applies to the main tone set, consonant set and all Pinyin input modes, and works in conjunction with the dual-mode tone function.

[0066] 22. The method described in technical solution 21, wherein the tone triggering method of single-key dual-phoneme combination is selectable: single tone in numerical order is triggered by numbers 1-4, 5; dual tone in numerical order is triggered by numbers 6, 7-0; and character tone is triggered by an independent key or a "punctuation-tone" shared key.

[0067] 23. The method described in technical solution 21 is characterized by supporting industry-specific extensions: medical, financial, and industrial fields can import exclusive dual-phoneme combination rule libraries, and the system automatically associates industry-specific tone combinations; the input response delay for professional terms with tone is ≤70ms, and the annotation accuracy is ≥99%.

[0068] 24. A multi-input mode human-computer interaction method based on the method described in technical solution 1 or 21, characterized in that it includes:

[0069] 1) Full coverage of input modes: It includes six input modes: abbreviated pinyin, abbreviated pinyin extension, superimposed pinyin, mobile double pinyin, single-key full pinyin, and double-key full pinyin. All modes are constructed based on the number sequence vowels described in technical solution 1.

[0070] 2) Core functions of each mode:

[0071] ① Simplified Pinyin: Supports encoding of Chinese and English initials / core syllables, bilingual mixed encoding, and index response latency ≤200ms;

[0072] ②Simplified Pinyin Extension: Superimposes the logic of "medial vowel + single-key number sequence vowel". The medial vowel can be triggered by either long press on the soft keyboard to pop up / directional sliding or long press on the hard keyboard to activate + short press to output.

[0073] ③ Overlay spelling: Triggered by the initial and medial vowels, compatible with abbreviated spelling, abbreviated spelling extension and full spelling modes, and supports mode conversion (accuracy ≥99%);

[0074] ④ Mobile Double Pinyin: Lowercase / uppercase initials and English characters are triggered by double keys; medial vowels with zero medial vowels are triggered directly by number-order vowels; the number-order dual-tone system in the number-order tone extension method of technical solution 7 and the single-key dual-phoneme combination rules of technical solution 21 are reused. After inputting an initial + number-order vowel, the system automatically triggers single-key dual-phoneme combination by using the 0 key (bound to medial i) and the "-" key (bound to medial u / ü), directly prompting the simplified code word or syllable symbol of the initial-medial-vowel-tone syllable containing the medial vowel. The combination response delay is ≤50ms (test environment: room temperature 25℃±5℃, device memory ≥2GB); the simplified code word or syllable symbol of the initial-medial-vowel-tone syllable without the medial vowel is supported by inputting the tone of the consonant set character in technical solution 10 (including independent tone keys or "punctuation-tone" shared key keys), or by reusing technical solutions. In Case 7, the numerical tone system input allows users to choose between two methods, sharing the tone-matched vocabulary mapping library. The 0 key corresponds to i + numerical vowel + tone, and the "-" key corresponds to u / ü + numerical vowel + tone. Furthermore, the mobile double-pinyin mode incorporates a splitting avoidance identification logic: for splitting behaviors such as "splitting the initial consonant input and the medial vowel trigger as two independent software calls" and "splitting the encoding generation and candidate prompts as different modules," it automatically identifies and rejects equivalent inputs by detecting the temporal correlation of the input signal (the interval between consecutive operations within the same input session is ≤100ms) and the module call chain. At the same time, it clarifies the splitting avoidance verification standard: if the input steps, encoding generation logic, and candidate output results of a certain scheme are ≥90% equivalent to the core features of this mode, and are achieved through splitting technology features, then it is determined to be an infringement.

[0075] ⑤ Single-key full pinyin input: Adapts to existing technology comparison scenarios, allowing users to input pinyin characters one by one;

[0076] ⑥ Double-key full spelling: Corresponds to the single-key full spelling logic of the letter keyboard, suitable for children's basic language teaching scenarios;

[0077] 3) Mode Coordination Switching: Supports automatic switching based on input sequence or manual switching by the user, with a switching delay of ≤50ms.

[0078] 25. The method described in technical solution 24, characterized in that the triggering rule of the soft keyboard of the simplified pinyin extension mode is: long press the 0 key / "-" key for ≥300ms (≥500ms in children's mode) to pop up a large candidate pop-up window, or "middle key + directional slide", with a recognition accuracy of ≥96%, and simultaneously display the single key dual phoneme combination of "middle key + tone".

[0079] 26. The method described in technical solution 24, characterized in that the mobile double-keyboard blind typing rules: support users to adjust the combination order; skilled users have a blind typing accuracy of ≥95%, support binding single-key dual-phoneme combination shortcut keys, and a response delay of ≤15ms.

[0080] 27. A multilingual human-computer interaction system based on the method described in any one of technical solutions 1-26, characterized in that it comprises:

[0081] 1) Input module: Receives user key operations, voice or gesture input, and outputs input signals with feature identifiers (key input signals include press duration / force features, and voice input signals include voiceprint / syllable features);

[0082] 2) Encoding Module: Pre-stores unmodifiable numerical vowel mapping rules and mother-child interface configurations, converts the input signal into corresponding initial-medial-final codes or multilingual character codes, and has a built-in encoding conflict detection unit. The rejection response delay for invalid encoding combinations (such as "zh+ü" or "b+iong") is ≤150ms (test environment: room temperature 25℃±5℃, humidity 40%-60%, no external electromagnetic interference); the numerical vowel mapping rules and mother-child interface configurations correspond to the methods described in any one of technical solutions 1-26, and are prevented from being tampered with by a hardware chip-level encryption unit;

[0083] 3) Candidate processing module: Based on the encoding, the encrypted offline dictionary is called to generate a candidate character / vocabulary list, which supports dual sorting by input frequency and semantic relevance, and the candidate list update delay is ≤30ms;

[0084] 4) Display module: Displays the vowel interface, candidate list and extended interface. The interface rendering follows the principle of "horizontal alignment between the trigger area and the candidate area", with an alignment error of ≤1 pixel. It supports adaptive screen size of multiple devices (adaptation range: 1.5 inches-21 inches).

[0085] 5) Control Module: Employing a multi-threaded collaborative architecture, the module controls the synchronous operation of the above modules, enabling input mode switching, multimodal interaction, and cross-device adaptation with a switching latency of ≤50ms. It includes a built-in anti-circumvention monitoring unit, focusing on scenarios such as technical feature splitting and device-specific circumvention. It identifies and intercepts such circumvention through the following rules: ① Monitoring module call chains and rejecting input behaviors that split "encoding conversion-conflict detection-candidate generation" into independent modules without collaborative verification; ② Detecting the temporal continuity of input signals and intercepting behaviors where splitting core operations (such as splitting key triggers and encoding mappings) is not a continuous operation; ③ Verifying the integrity of core technical features and directly rejecting and recording circumvention behaviors that only call some modules (such as only calling the input module and display module, skipping the encryption verification of the encoding module) to achieve equivalent input; ④ For small-screen devices such as smartwatches, intercepting circumvention behaviors such as "using only '11' and '12' extended keys but deviating from the core mapping table" and "modifying the key-vowel correspondence of small-screen devices."

[0086] 6) Hardware encryption unit: Integrated with the encoding module, it performs hardware-level encryption storage of core mapping rules and offline dictionaries, allowing reading only through authorized interfaces and prohibiting modification; the encryption algorithm adopts the Chinese national standard SM4 and the international standard AES-256, supports automatic switching of encryption standards according to the compliance requirements of the target country, and generates encryption compliance reports that conform to local standards, adapting to the encryption compliance requirements of different countries.

[0087] 28. The system as described in technical solution 27, characterized in that it further includes an intelligent switching module: automatically switching the input mode based on the user input sequence or multimodal interaction features; entering the overlapping compatible simplified pinyin mode when inputting "single / double-key initials"; switching to the simplified pinyin extended mode when inputting "initial + medial"; entering the overlapping dedicated mode when inputting "initial, medial, and double-key"; switching to the corresponding tone mode when triggering tone input; switching to the corresponding interaction mode when triggering multimodal input; and automatically switching to the corresponding mode when a single-key dual-phoneme encoding sequence is detected; the switching delay is ≤50ms.

[0088] 29. The system as described in technical solution 27 is characterized in that it further includes an AI collaboration module: interacting with a generative AI large model, realizing semantic-level input completion and multi-turn dialogue encoding collaboration functions based on numerical vowel encoding logic, with an AI collaboration accuracy of ≥92%; recommending high-frequency single-key dual-phoneme combinations based on input context, with a word completion accuracy of ≥92%.

[0089] 30. The system as described in technical solution 27 is characterized by supporting Android 8.0+, iOS 12.0+, Windows 10+, Linux (Ubuntu 18.04+), and industrial real-time operating systems (VxWorks 7.0+, QNX 7.0+); cross-platform compatibility ≥99%, installation package size: basic version ≤20MB, full version ≤50MB; and for different countries' patent examination requirements, it supports compliance adaptation of the configuration parameters of the encoding module and encryption unit, meeting the data security and encryption standards of various countries without changing the core technical solution; furthermore, the system has a patent infringement early warning function, which can monitor and record counterfeit use of core encoding rules and parent-child interface logic, and generate an infringement evidence chain containing timestamps and device identifiers; a new technical feature splitting and circumvention verification scheme is added, specifically including: 1) Test scenario: simulating the splitting of the "three-in-one mapping table" of technical solution 1 (only implementing vowel-key mapping, skipping encoding sequence binding), splitting the " 1) Single-key dual-phoneme triggering logic (first triggering the key and then generating dual-phoneme encoding through a third-party module), and the "module collaboration logic" of splitting technical solution 27 (only calling some modules to achieve equivalent input); 2) Test method: Under a standard test environment (room temperature 25℃±5℃, humidity 40%-60%, device memory ≥2GB), the above splitting avoidance input behavior is generated through an automated test tool, and the system response results are recorded; 3) Judgment criteria: If the system can 100% identify and intercept the above splitting avoidance behavior and generate a warning log containing "splitting behavior type, trigger time, device identifier", it is judged as passing the verification; if there is uninterrupted splitting avoidance behavior, it is judged that the core technical features are not effectively protected, and anti-avoidance logic optimization needs to be started.

[0090] Detailed descriptions of the attached figures and tables:

[0091] The most significant invention of this application is the number-sequential vowels and tone-marking medials. The attached diagrams, which can be annotated, all indicate the corresponding number-sequential vowels and the same tone-marking medials, which are 1ī / ū&ǖ, 2í / ú&ǘ, 3ǐ / ǔ&ǚ, 4ì / ù&ǜ, 5i / u&ü, 6ù&ǜ / ì, 7ū&ǖ / ī, 8ú&ǘ / í, 9ǔ&ǚ / ǐ, 0(ī), -(ū&ǖ). This means that when 0 is the guiding key (ī), then 1 is ī, 2 is í, 3 is ǐ, 4 is ì, and 5 is i, forming a set of number-sequential single-tone tone-marking medials. And when "-" is the guiding key (ū&ǖ), then T1 (corresponding to 6) is ù&ǜ, T2 (corresponding to 7) is ū&ǖ, T3 (corresponding to 8) is ú&ǘ, and T4 (corresponding to 9) is ǔ&ǚ, forming a set of character tone-marking medials.

[0092] Conversely, if 0 is the key (ū&ǖ), then 1 is ū&ǖ, 2 is ú&ǘ, 3 is ǔ&ǚ, 4 is ù&ǜ, and 5 is u&ü, forming a set of tone marks for single-tone numbers. And if "-" is the key (ī), then T1 (corresponding to 6) is ì, T2 (corresponding to 7) is ī, T3 (corresponding to 8) is í, and T4 (corresponding to 9) is ǐ, also forming a set of tone marks for character tones.

[0093] Strictly speaking, not only 9 and "-", but any combination of two numeric keys can be designated as the guide key for tone markers. It's just that 9 and "-" are less frequently used in Chinese, Chinese-English bilingual, or Chinese-English and multilingual modes. The software system can be configured according to various user needs and actual usage effects.

[0094] This application involves seven types of hardware and software keyboards and their extended interfaces, as well as various derivative types of hardware and software keyboards and their extended interfaces, including the four conventional types shown in Figures 1-4 and the three unconventional types shown in Figure 42 (left-angled, right-angled, and top-positioned). These can also form mutually cooperating hardware and software keyboards and their extended interfaces. Furthermore, they can be mutually cooperated with the hardware and software letter keyboard and its interface shown in Figure 44, or the three multi-directional hardware and software letter keyboards and their interfaces shown in Figure 42 (left-angled, right-angled, and top-positioned), to form mutually cooperating hardware and software large, medium, and small keyboards and their interfaces.

[0095] Numerical symbols and extended characters form the basis of the character layout in this application. The character layouts in Figures 1-4 are essentially identical. Figure 1 shows a 4x4 independent hard and soft keypad and its extended interface, labeled with numerical symbols and extended numerical symbols, English characters, initials, medials, and finals. The finals are arranged in numerical order to form numbered finals. Ignoring the tone keys in the entire column to its right, this is a 4x3 independent hard and soft keypad and its extended interface. Figure 2 shows a smartwatch's hard and soft keypad and its extended interface, with 1-12 numerical symbols ordered clockwise. In the middle are four tone keys that share keys with punctuation marks. Figure 3 shows an independent single-row hard and soft keypad and its extended interface for occasional human-computer interaction, suitable for narrow, elongated devices. Figure 3 can also be entirely transferred to the numerical row of Figure 4, thus Figure 4 can be viewed as an embedded 16-key hard and soft keypad and its extended interface, expanded by adding four specified tone keys to the symbol keys. Figure 4 shows two hard and soft keyboards embedded in a regular keyboard, arranged in ascending and descending order, along with their extended interfaces. Ignoring the tone keys on the left and right, the corresponding 4x3 and 3x4 embedded hard and soft keyboards and their extended interfaces are shown.

[0096] Figures 5, 8, 11, 14, 17, 20, 23, 26, 29, and 32 show the keyboard layouts for a 4x4 hard / soft keypad and its extended interface, including numeric characters, extended numeric characters, Chinese-English characters, ordinal numbers, and pinyin examples. If the tone keys on the right are ignored, the layouts are for a 4x3 hard / soft keypad and its extended interface.

[0097] Corresponding to the types in Figures 1-4, Figure 5 shows the independent hard and soft keyboards and their extended interfaces. In addition to marking letters and initials, it also marks example characters for initials and finals. If the example characters of adjacent initials and finals can be combined into one example character to replace the two, they are combined into one example character. The same applies below. Figure 6 is a schematic diagram of the initial-medial input method of Figure 5, which uses the second medial vowel to be the same first and then change. It shows the input of any initial-medial vowel and example character or a choice between the two using a two-key combination. The bottom row is the specific explanation of the second medial vowel according to the position of the initials. The top row is the numbered finals and the numbered medial vowels with tone marks. Figure 7 is a schematic diagram of the initial-medial input method of Figure 5, which uses the initial-medial input method of the same first and then change. It shows the input of any initial-medial vowel and example character or a choice between the two using a two-key combination. The leftmost column of Figures 6 and 7 is the first key of the key. The top row of each key is the second key to be pressed. The bottom row of each key is the explanation of the initial-medial or initial input method of the first, second or last character.

[0098] Similar to Figures 5-7, Figures 8-10, 11-13, 14-16, 17-19, 20-22, 23-25, 26-28, 29-31, and 32-34, which are adapted to various types of Figures 1-4, all feature an independent hardware / soft keyboard and its extended interface as the first image in each group. Besides labeling letters and initials, example characters for both initials and finals are also provided. The subsequent images in each group illustrate the use of the initial-medial input method (first the same initial, then changing) for the second medial vowel, allowing input of any initial-medial vowel and an example character or a choice between the two using a two-key combination. The final image in each group illustrates the use of the initial-medial input method (first the same initial, then changing) for the first image, allowing input of any initial-medial vowel and an example character or a choice between the two using a two-key combination. In the sub-figures and final figures of each group of figures, the leftmost column of each figure represents the first key of the button, the top row of each figure represents the second key to be pressed, and the bottom row of each figure represents the specific input method of the initial consonant, second consonant, or final consonant.

[0099] The above are examples of various hardware and software keypads and their extended interfaces. Many similar examples can be deduced by analogy, all of which are extensions and expansions of this application. The following mainly describes the vowel layout of various hardware and software keypads and their extended interfaces.

[0100] The vowel layouts and numerical order of vowels in Figures 5-7, 8-9, and 10-12 are different. The vowel layouts and numerical order of vowels in the remaining groups of figures are consistent with those in Figure 8-9. The corresponding keyboard layouts for the other groups of figures in Figures 5-7 and 10-12 can be derived from Figure 8-9, and will not be further explained.

[0101] Figures 35-37 are tables showing the correspondence between initial consonants and medial vowels, and their combinations in three-part combinations (i.e., overlapping initial consonants and medial vowels). The top row contains the medial vowels, and the left column contains the initial consonants. Similarly, by arbitrarily adjusting the character layout and matching vowel layout, rhyme list, and medial vowels of any hard or soft keyboard and its extended interface, more similar tables can be derived, which will not be further explained. In Figures 35-37, Figure 35 shows the correspondence between the initial consonant and medial vowel combination formed by inputting any initial consonant and example character or one of the two keys in Figures 5-7, and the medial vowel combination formed by the third key. Figure 37 shows the correspondence between the initial consonant and medial vowel combination formed by inputting any initial consonant and medial vowel and example character or one of the two keys in Figures 10-12, and the medial vowel combination formed by the third key. Figure 36 shows all the other groups of figures, namely Figures 8-9, 11-13, 14-16, 17-19, 20-22, 23-25, 26-28, 29-31, and 32-34. The first two keys are used to input any initial consonant and example character or one of the two, and the third key is used to input the numbered finals to form the initial consonant-final correspondence table.

[0102] In Figures 35-37, the vowel uo within square brackets represents a conventional initial-final combination, leading to duplicate initial-medial vowels and syllables for the same initial consonant u and uo. This can be avoided by not splitting the vowel uo and applying it uniformly as o, or by treating uo as o before combining it with initials other than b, p, m, and f, and then restoring it to uo after combination. The three-part initial-medial vowel tables in Figures 35-37 are quite complex, all derived from the keyboard diagram and double-part initial-medial vowel tables in Background Technology 1. The difference between the double-part initial-medial vowel tables of the numeric keypad and the main keyboard is that the numeric keypad's initial-medial vowel is a three-part combination of an initial consonant and a medial vowel, which is then combined with a vowel. The main keyboard's initial-medial vowel is a double-part combination of an initial consonant and a medial vowel. The initial-medial vowel is the same, but the former's three-part combination is more complex, while the latter's double-part combination is simpler. This necessitates compatibility between the numeric keypad's three-part combination and the main keyboard's double-part combination, but existing technologies are incompatible. The double-pinyin scheme of the Hanyu Pinyin system had to be abolished because the simplified symbols were difficult to split or combine in three-part spelling. The original symbols were restored to make it so that all medial vowel symbols could be easily split or combined in double or three-part spelling. This improved the scheme into a digital version of the Hanyu Pinyin system that is compatible with both double and three-part spelling.

[0103] Figures 38-41 are examples of candidate characters for the first key prompt in the simplified pinyin input mode, which are the basic simplified pinyin modes for Chinese and / or English.

[0104] Figure 42 shows the left / right / top layout of the double-pinyin keyboard and its embedded single-row, numeric keypad, and smartwatch.

[0105] Figure 43 is a functional partition diagram of the generalized dynamic extension interface, including, from top to bottom, the associated prompt area 35% + dynamic candidate area 30% + fault-tolerant prompt area 20% + multimodal interaction area 10%.

[0106] Figure 44 is a keyboard diagram of the initials and medial vowels in Background Technology 2, which is the same as the numbered vowels in Figures 1-4 and can form a mother-child interface.

[0107] Figure 45: Cross-device adaptation layout comparison diagram, corresponding to the cross-device key surface size rules of smartphones, watches and industrial panels;

[0108] Figure 46: Flowchart of switching between multiple input modes including simplified pinyin, mobile double pinyin, simplified pinyin extension, overlapping pinyin, and full pinyin index;

[0109] Figure 47 is a diagram of the core support engine architecture, showing the flow of data verification in the basic layer, pattern parsing in the core layer, dictionary matching in the support layer, result conversion in the output layer, dictionary matching feedback, input data after verification, and dictionary data matching after parsing.

[0110] Figure 48 is a hierarchical architecture diagram of a multimodal human-computer interaction system, showing the data flow of input from the hardware layer's input devices (soft and hard keyboards / extended interfaces) and the software's driver software, as well as the interaction commands and feedback signals from the hardware layer's output devices (touchscreens, displays, audio and video devices).

[0111] Figures 49-52 are keyboard prompts for the mobile double-pinyin input method, showing the lowercase and uppercase initials and letters, number vowels and number tone markers, and the tone markers (ī) and (ū & ǖ) following 0 and "-".

[0112] Table 1: Phonetic alphabetical characters for beginners with the fewest characters and their syllable examples, suitable for children's teaching scenarios and adapted to children's teaching;

[0113] Table 2: Examples of syllable abbreviations for initial-medial-final-addition-number-ordered double tone, corresponding to number-ordered double tone;

[0114] Table 3: Examples of simplified code characters for numerically ordered initials and tones of vowels with additional initials and tones, corresponding to the consonant set (initials and tones medial vowels);

[0115] Table 4: Examples of numerical code characters for initial and medial vowels with additional numerical order, corresponding to the consonant set (initial, medial, tone, and vowel input);

[0116] Table 5: Graphical user interface diagram of the teaching interface of the sound, word, sentence and human-computer interaction method, and corresponding extended interface.

[0117] Beneficial effects of each technical solution

[0118] 1. Construct an independent number sequence vowel encoding system to achieve seamless synchronization of key mapping across devices, enabling multi-language input without switching modes. The core technical feature is indivisibility; temporal continuity and encoding integrity detection prevent splitting and avoidance, ensuring robust protection and adapting to efficient input needs across multiple scenarios such as mobile terminals, smart wearables, and industrial control equipment.

[0119] 2. Offers 10-16 key interface configurations to meet the minimalist layout requirements of small-screen devices, while also reserving space for multi-language and tone input by expanding the key layout, adapting to different device hardware and usage scenarios.

[0120] 3. Clarify the rules for splitting / merging vowels, and achieve flexible adaptation while keeping the core encoding logic unchanged, so as to ensure the coordination and consistency of different groups of vowels and improve the versatility of the solution.

[0121] 4. Offers three sorting schemes: efficient, alphabetically compatible, and children's educational, to suit users who frequently input text, are accustomed to alphabetical keyboards, and children's learning scenarios, covering the needs of different user groups.

[0122] 5. The dynamic keyboard is displayed in layers according to screen size, which reasonably controls the screen occupancy ratio and takes into account both the ease of operation on small screen devices and the functional expandability on large screen devices.

[0123] 6. Enables automatic adaptation between dynamic keyboards and physical keyboards, adaptive touch area for small-screen devices, and optimized error-touch filtering, significantly improving touch recognition accuracy and reducing discomfort when inputting across devices.

[0124] 7. An innovative single / dual tone system is introduced, which supports the simultaneous prompting of two sets of tone candidate syllables and dynamically adjusts the number of candidates according to the screen size to meet different tone input needs and visibility requirements.

[0125] 8. Supports priority binding of high-frequency tones, optimizes tone key configuration for adult and children's scenarios, reduces tone input error rate, and improves input efficiency for different user groups.

[0126] 9. The number of candidate prompts and the size of characters can be precisely adjusted according to the screen size, and page switching can be switched by sliding, balancing the display clarity of small screen devices and the input efficiency of large screen devices.

[0127] 10. Offers dual configurations of independent tone keys and "punctuation-tone" shared keys, allowing for seamless expansion of the consonant set without switching modes, enriching tone input methods and improving input flexibility.

[0128] 11. The "Punctuation-Tone" shared key distinguishes functions through light touch / long press, supports custom duration thresholds, enables reuse of punctuation and tone input, simplifies interface layout, and improves operational efficiency.

[0129] 12. The consonant set is automatically activated in specific scenarios, and the pinyin splitting animation is synchronized in children's teaching scenarios to improve input fluency and teaching intuitiveness.

[0130] 13. Supports multiple activation methods such as shortcut keys, voice, and gestures, is compatible with multimodal input and output, and the interface can be dragged and scaled to adapt to different display modes, meeting diverse operating habits and device adaptation needs.

[0131] 14. The error-tolerant prompt area calculates the probability of accidental touch based on the physical distance of the button and the similarity of the voice. The Top 1 correction candidate has a high accuracy rate, effectively reducing the mis-input rate and improving the input experience.

[0132] 15. The multimodal interaction area optimizes the recognition logic for adults, children, and single-key dual-phoneme collaborative scenarios, achieving accurate recognition of professional terms, non-standard pronunciations, and syllables with tones.

[0133] 16. Cross-device differentiated configuration of touch areas, multi-directional layout for adaptive document typesetting, scenario-based customization to meet the needs of teaching, industry, and cross-border communication, and the core engine provides offline verification, parsing, and translation support to ensure input reliability in multiple scenarios.

[0134] 17. The children's teaching scenario integrates pronunciation and mouth shape animation, multiple speech speed adjustment and interactive reading games, and visualizes the decomposition of single-key double phoneme combinations, enhancing the fun and effectiveness of teaching and helping children learn pinyin.

[0135] 18. The offline encoding verification module quickly rejects invalid initial-medial-initial combinations with low response latency. When a failure occurs, it provides a list of recommended combinations to guide the user to enter the correct information and improve input accuracy.

[0136] 19. Supports bidirectional offline translation between Chinese and English and languages ​​with tones. The translation accuracy of commonly used phrases and professional terms is high. Single-key dual-phoneme rules are reused in languages ​​with tones, reducing the conflict rate of bilingual mixed input.

[0137] 20. In industrial control scenarios, the core instructions are cached offline in large quantities and for a long period of time. The input time for instructions with adjustment is short and the availability is high when the network is disconnected, meeting the needs of efficient and stable input in industrial scenarios.

[0138] 21. Provides two irreplaceable single-key dual-phoneme triggering methods: long press and shortcut key combination. System permission locking prevents third-party access, and trigger response is fast and conflict-free. It explicitly states that "single-key triggering - encoding mapping - cross-mode compatibility" are inseparable, preventing circumvention behaviors related to split triggering and encoding generation.

[0139] 22. Supports multiple triggering modes for number sequence single / double tone and character tone, adapting to the input habits of different users and improving the flexibility of using single-key dual-phoneme combinations.

[0140] 23. Supports importing dual-phoneme rule bases specific to industries such as medical, financial, and industrial. Professional terminology input with tone has low response latency and high annotation accuracy, adapting to the precise input needs of professional scenarios.

[0141] 24. It covers six input modes, including simplified pinyin, overlapping pinyin, and mobile double pinyin, and automatically switches based on the input sequence. The mobile double pinyin mode has built-in splitting avoidance recognition logic. It prevents infringement through temporal correlation and module link detection, meeting the needs of different input scenarios and habits.

[0142] 25. The simplified pinyin extension mode can be triggered by long-pressing the pop-up window or directional swiping. The large candidate pop-up window improves the ease of operation, has a high recognition accuracy, and ensures accurate and efficient input of pinyin.

[0143] 26. The mobile dual-keyboard allows users to adjust the combination order, and skilled users have a high accuracy rate in blind typing. The shortcut key binding has a low response latency, meeting the needs of efficient blind typing input.

[0144] 27. All modules work together to achieve efficient operation of the entire process from input, encoding, candidate generation, to display. Hardware chip-level encryption ensures the core mapping rules and dictionary are tamper-proof. A built-in anti-circumvention monitoring unit intercepts circumvention behaviors such as module splitting and device-specific customization. Dual encryption standards adapt to global compliance requirements.

[0145] 28. Automatically switch the corresponding input mode based on the input sequence and multimodal features. The switching latency is low, no manual operation is required from the user, ensuring a smooth and continuous input process and improving the user experience.

[0146] 29. It interacts with generative AI large models to achieve semantic-level input completion and multi-turn dialogue encoding collaboration, accurately recommends high-frequency single-key dual-phoneme combinations, and has high accuracy in word completion with tone, greatly improving input efficiency.

[0147] 30. Compatible with multiple platforms including Android, iOS, Windows, and industrial real-time operating systems, with high cross-platform compatibility and a moderate installation package size. It features patent infringement warning capabilities, generates a complete chain of evidence, and ensures effective protection of core technical features by breaking down verification schemes, adapting to data security and patent examination requirements in various countries. Detailed Implementation

[0148] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0149] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0150] Traditional input method technologies suffer from structural defects in multi-device adaptation, multi-language compatibility, multi-scenario coverage, and multimodal collaboration. Keyboard layouts and input logic are fragmented across devices, leading to high accidental touch rates on small-screen devices; tone marking methods are isolated, lacking a unified mapping for tones across multiple languages; multimodal interaction response is delayed and logically disconnected; core input functions and extensions are forcibly bound, making them easily circumvented and inefficient; multi-language encoding systems are independent, causing conflicts in bilingual input; there is a lack of parent-child interface collaboration, resulting in low efficiency across devices; and multiple input modes are fragmented, making switching inconvenient and inefficient for initial consonant input. These problems lead to a fragmented input experience, high learning costs, and insufficient scenario coverage, stemming from the fact that existing technologies have not fully considered the phonetic rules of Chinese Pinyin.

[0151] In response, this application proposes a multilingual human-computer interaction method based on a number-sequence vowel interface. By defining a unique set of number-sequence vowels and mapping rules, configuring cross-device compatible digital key positions and extended key position systems, an irreplaceable number-sequence vowel interface is constructed, a collaborative parent-child interface system is built, and a unified multilingual encoding logic is implemented. This forms a core technology system that does not rely on existing English keyboards and has an independent number-sequence vowel encoding system, effectively solving many problems existing in the prior art and improving the efficiency and universality of human-computer interaction.

[0152] For ease of understanding, the following explains some key terms in this embodiment:

[0153] Numbered finals refer to a set of finals formed by splitting or merging medial finals in Chinese Pinyin according to the principle of "consistency of tongue position in pronunciation". Each group of finals in this set is assigned a unique digital key position, which, together with the key position identifier and the encoding sequence, forms a three-in-one mapping relationship, aiming to optimize input efficiency and logic.

[0154] The principle of "consistent tongue position during pronunciation" means that when constructing a set of numbered vowels, vowels with similar or identical tongue positions during pronunciation are grouped together to reduce the cognitive burden and operational complexity for users during input and improve the intuitiveness of input.

[0155] The "vowel-key identifier-encoding sequence" three-in-one mapping table is a data structure that uniquely binds a numerical vowel, its corresponding key identifier, and the generated encoding sequence. This mapping table is the core of the system, and its contents are designed to be unmodifiable through conventional software configuration to ensure the stability and security of the input logic.

[0156] Extended key identifiers 11 and 12 are additional key identifiers introduced to adapt to small-screen smart wearable devices such as smartwatches. These identifiers, along with the core digital keys 1-0, are incorporated into a three-in-one mapping table of "vowel-key identifier-encoding sequence" to enable single-key triggering of more vowels within limited screen space.

[0157] USB HID and Bluetooth HID protocols are communication protocols used to achieve seamless synchronization of number sequence vowel mapping rules and candidate prompt sequences between different devices (such as mobile terminals, smart wearable devices, and industrial control equipment). These protocols ensure a consistent input experience across devices.

[0158] The "high-frequency vowel proximity principle" refers to assigning frequently used vowels to keys that are easier for users' fingers to reach in the key layout design of the numerical vowel interface, in order to improve input speed and ease of operation.

[0159] The "minimalist layout principle for small-screen devices" refers to adopting a simplified key layout for devices with smaller screen sizes, such as smartwatches. Prioritize the use of extended key markers 11 and 12 to achieve single-key triggering of core vowels, in order to minimize accidental touches and optimize the user experience.

[0160] The parent-child interface system refers to a collaborative system consisting of a keyboard with the same numerical vowel and consistent numerical string layout and sorting, along with its extended interfaces. This system shares the same encoding mapping table and candidate word library, ensuring consistency in input logic and candidate lists across different devices or interface formats, and achieving latency-free switching through a control module.

[0161] The reserved coding space for vowels refers to the coding space reserved in the core coding rules for vowels or phonetic features of different languages. By combining phonetic feature mapping, unified coding for multiple languages ​​can be achieved, avoiding language mode switching.

[0162] Speech feature mapping refers to the process of mapping speech features (such as phonemes and tones) of different languages ​​to reserved coding bits for numbered vowels. This enables the system to process input from multiple languages ​​based on a unified coding logic.

[0163] The core distinguishing and indivisible feature refers to the key difference between this method and existing letter keyboard input methods, namely, the implementation of input through an independent numerical vowel encoding system. This feature emphasizes the integrity and synergy of the various technical features in the method; any attempt to separate or implement only some features is considered to fall within the protection scope of this method.

[0164] The standard and technical logic for identifying and deconstructing circumvention methods refer to the standards and methods used to detect and determine whether there are attempts to bypass the core logic of this method to achieve equivalent input behavior by splitting the core functional links of this method (such as vowel mapping, key configuration, interface construction, collaborative interaction, and encoding logic) or modifying the core mapping relationships. This logic identifies circumvention behavior by detecting the temporal continuity of input operations and the integrity of the encoding sequence.

[0165] This embodiment provides a multilingual human-computer interaction method based on a number sequence vowel interface.

[0166] First, a unique set of numerically ordered finals and their mapping rules are defined. This set of numerically ordered finals is formed by manually splitting or merging medial finals in Pinyin according to the principle of tongue position consistency. For example, finals with similar tongue positions, such as a, ang, eng, ang, en, en, e / üe / er, u / o / uo / i, ao, ei / ü, ou, and ai, can be manually grouped together, forming 11 core finals. Then, a unique numeric keypad is assigned to each group of finals; for example, the first group is assigned to key 1, the second to key 2, and so on. Based on this, a mapping table is created that statically binds the numerically ordered finals, their corresponding keypad identifiers, and the generated encoding sequences. This mapping table is fixed during system initialization and does not provide users with the option to modify it through the regular software interface, ensuring its stability and security.

[0167] Secondly, a cross-device compatible numeric keypad and extended keypad system are configured. Based on the core numeric keypad 1-0, extended keypad identifiers 11 and 12 are introduced, corresponding to extended characters such as "-" and "=". These extended keypad identifiers, along with the core numeric keypad, are incorporated into the aforementioned fixed "vowel-keypad identifier-encoding sequence" three-in-one mapping table. When the system runs on different devices, such as mobile terminals, smart wearable devices, or industrial control equipment, these devices are configured to share the same set of numeric vowel mapping rules and candidate prompt sequences. Synchronization of the mapping rules can be achieved by manually copying the mapping table file to each device or by distributing it through a simple network service, ensuring that all devices use the same input logic.

[0168] Furthermore, an irreplaceable numerical vowel interface is constructed. This interface is designed to contain a fixed number of keys, such as 12 keys, including core numeric keys and extended keys 11 and 12. The key layout follows the principle of placing high-frequency vowels close to the hand, for example, placing the most frequently used vowels in the central area of ​​the keyboard. It also adheres to the principle of minimalist layout for small-screen devices; for example, on small-screen devices such as smartwatches, only the most essential keys are displayed, and extended keys 11 and 12 are used to trigger more core vowels, enabling single-key input of vowels. The layout of this interface and the numerical vowel mapping rules are designed to be strongly bound, preventing users from modifying the correspondence between core keys and vowels through conventional interface customization functions. In addition, the key layout of small-screen devices such as smartwatches is configured to be synchronized with the parent-child interface system to ensure consistent input logic when switching between different devices.

[0169] Therefore, a collaborative parent-child interface system is constructed. This system consists of a main keyboard interface with identical numerical vowel sequences and consistent numerical string layout and sorting, along with its extended interfaces, such as a full-size soft keyboard interface and a simplified interface designed for smartwatches. These interfaces share the same encoding mapping table and candidate word library, ensuring that the system outputs a consistent candidate list when the user inputs the same initial-medial combination on different interfaces. A control module is configured to manage the switching between the parent and child interfaces, for example, by detecting the currently active device type or user action, and ensuring that the current input context is not interrupted during the switching process.

[0170] Furthermore, a unified encoding logic for multiple languages ​​is implemented. Based on the aforementioned core encoding rules for number-order vowels, the system reserves specific encoding positions for vowels or phonetic features of different languages ​​and combines this with phonetic feature mapping to achieve unified encoding for Chinese, English, and other languages ​​with tones. For example, a language identifier can be added before or after the vowel encoding, or the phonetic features of a specific language can be directly mapped to the reserved encoding positions. Thus, users can complete bilingual mixed input on the same input interface without manually switching language modes, and the encoding sequences for different languages ​​are designed to be mutually non-conflicting.

[0171] Finally, this method possesses core technological distinctions and indivisible characteristics. It implements input by constructing an independent numerical vowel encoding system, independent of the character mapping relationships of existing letter keyboards, thus significantly differentiating itself from existing input methods based on letter key reuse. The aforementioned technical features (1)-5)—defining the vowel set, configuring key positions, constructing the interface, collaborative linkage, and unified encoding—are designed as an indivisible whole to realize the core functionality of this method; they require collaborative linkage to function. Any attempt to break it down into independent steps or modules, or to implement only some features, is considered to fall within the protection scope of this method. Furthermore, to identify and prevent splitting circumvention behavior, the system is configured to detect the temporal continuity of input operations, such as determining whether the time interval between consecutive key presses is too long, and the integrity of the encoding sequence, such as checking whether any link in the three-in-one mapping table is missing, to identify potential splitting circumvention behavior. The numerical vowel encoding system of this method is considered a unified core framework; all subsequent cross-device adaptations, scenario expansions, or mode optimizations must be implemented based on this framework, and independent input logic is not allowed to be built outside of this framework.

[0172] This method achieves unified and seamless synchronization of cross-device input logic by defining a unique set of numerical vowels and a three-in-one mapping table, effectively solving the fragmentation problem of traditional input methods in multi-device adaptation. Its irreplaceable vowel interface design, combined with the principle of proximity to hands for high-frequency vowels and a minimalist layout for small screens, significantly reduces the accidental touch rate on small-screen devices and improves input efficiency. The collaborative parent-child interface system and unified multilingual encoding logic enable bilingual mixed input without mode switching and without conflicts between different language encodings, greatly optimizing the multilingual interactive experience. This independent encoding system fundamentally overcomes the shortcomings of unreasonable phoneme layout and low input efficiency in existing technologies, providing users with a coherent, efficient, and universally applicable human-computer interaction solution.

[0173] In some of the embodiments described above in this application, a multilingual human-computer interaction method based on a number-sequential vowel interface is proposed. This method achieves an independent input system independent of existing English keyboards by defining a unique set of number-sequential vowels and mapping rules, configuring cross-device compatible numeric keys, constructing an irreplaceable number-sequential vowel interface, a collaborative parent-child interface system, and unified multilingual encoding logic. However, in practical applications, a single interface configuration may be insufficient to fully adapt to different device sizes, user habits, and diverse language input needs. For example, on very small screen devices, too many keys may lead to accidental touches, while in scenarios requiring frequent input of accented languages ​​or mixed multilingual input, the lack of dedicated interface support will reduce input efficiency.

[0174] In this regard, this application further proposes a specific configuration for the numerical vowel interface, which may include any of the following types: 11-key basic vowel interface, 10-key compressed vowel interface, 12-key multilingual vowel interface, 15-key multilingual vowel interface, or 16-key full-function character layout.

[0175] Specifically, the 11-key basic vowel interface is a basic and universal vowel interface configuration designed to provide direct input of core vowels while supporting some commonly used symbols or functions through extended keys. This interface consists of 10 core numeric keys (1-9, 0 (corresponding to 10)) and one key (marked as 11) corresponding to the extended character "-". These keys are bound to 11 groups of core vowels and extended characters according to a preset "vowel-key identifier-encoding sequence" mapping table. For example, key 1 might be bound to the high-frequency vowel "ɑ", key 2 to "ɑn", and so on, while key 11 is used to input the "-" character or as a trigger key for specific functions. This configuration ensures rapid input of core vowels while providing scalability for basic functions.

[0176] The 10-key compressed vowel interface is designed for scenarios with higher requirements for interface space or those pursuing ultimate simplicity. It reduces the number of keys by merging vowel groups, thus achieving a more compact layout. Building upon the 11-key basic vowel interface, adjacent groups of vowels are logically merged, allowing them to share a single key. For example, the vowels "ei / ü" and "ou," originally bound to two different keys, can be merged into the same key, distinguished by short / long presses or swipes, or intelligently determined based on context. Ultimately, the interface consists of only 10 keys: 1-9 and 0, further simplifying the interface and making it particularly suitable for very small screen devices or scenarios requiring one-handed operation.

[0177] The 12-key multilingual vowel interface adds a dedicated extended key to the basic vowel interface to reserve space for multilingual encoding, thus better supporting mixed multilingual input. Based on the 11-key basic vowel interface, a new key (marked as 12) corresponding to the extended character "=" is added. This newly added 12-key can be used as a multilingual switching key, a specific language character input key, or combined with the core vowels to trigger phonemes or characters of a specific language. For example, when inputting English, the 12-key can be used to input letters or symbols not commonly found in Chinese Pinyin, or as a trigger key to switch to English word input mode. This configuration provides a clearer entry point and expanded capabilities for multilingual input.

[0178] The 15-key multilingual vowel interface is designed to directly support tone input in languages ​​with tones (such as Chinese). By adding dedicated tone keys, it improves the efficiency and accuracy of tone input. Based on the 11-key basic vowel interface, four tone keys (T1-T4) are added. These keys can be directly used to input the four tones of Chinese (high level, rising, falling-rising, and falling), or one key can correspond to one tone. Another implementation method is to use shared "punctuation-tone" keys, where certain punctuation keys can be switched to tone input under specific operations (such as long press). For example, a light touch on the comma key inputs a comma, while a long press inputs the first tone. This configuration allows users to conveniently input syllables with tones without switching input modes.

[0179] The 16-key full-function character layout combines multilingual support and tone input functionality, providing a comprehensive and flexible interface layout to meet complex input needs. Based on the 12-key multilingual vowel interface, four additional tone keys (T1-T4) are added, or a shared "punctuation-tone" key layout is adopted. This means that the interface not only includes core vowel keys and multilingual extended keys, but also integrates tone input functionality. For example, users can complete Chinese Pinyin input with tone, English word input, and common symbol input on a single interface without frequently switching interfaces or modes. This layout offers maximum functional integration and is suitable for professional users who need to handle multiple languages ​​and complex text input.

[0180] By providing various specific configurations of the numerical vowel interface, this application effectively solves the problem of insufficient interface adaptability of the basic solution under different device sizes, user habits, and language input requirements. For example, for small-screen devices, the 10-key compressed vowel interface significantly reduces the number of keys by merging vowel groups, enabling efficient input even in limited screen space and avoiding accidental touches due to small keys. Meanwhile, the 12-key, 15-key, and 16-key multilingual vowel interfaces, by reserving or directly integrating multilingual encoding space and tone keys, allow users to perform multilingual mixed input and tone-based input without switching input modes, greatly improving input efficiency and convenience. Especially when dealing with toned languages ​​such as Chinese, accurate syllables can be directly input, reducing subsequent proofreading and correction work. These diverse interface configurations enable the numerical vowel interface to flexibly adapt to various application scenarios, from minimalist to full-featured, ensuring consistency of input logic across devices and optimizing user experience.

[0181] In some of the embodiments described above in this application, although a multilingual human-computer interaction method based on a number-order vowel interface is proposed and 11 core vowels are defined, detailed guidance has not yet been provided regarding the specific construction method of the number-order vowel set, particularly how to efficiently handle complex medial vowels, how to balance the types of vowels with the flexibility of key mapping, and how to adapt to different devices and user needs. This may lead to a lack of clear optimization strategies in the construction of the vowel set in practical applications, affecting its adaptability and user experience in different input scenarios.

[0182] In response, this application further proposes a specific method for constructing the aforementioned set of numerically ordered finals, including splitting the medial finals other than uo and üe into medial vowels and finals, and merging them into 11 sets of adaptable finals (ɑ, ɑnɡ, engɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi). Among these, ɑ and u / o / uo / i are in fixed positions, en only interchanges with e / üe / er, and the remaining finals can be interchanged arbitrarily; when there is no need for adaptation sharing, the position of each final is unrestricted; and two sets of finals can be merged into 10 sets of compressed numerically ordered finals.

[0183] Specifically, the method splits medial vowels other than uo and üe into medial vowels and final vowels, aiming to simplify the types of final vowels and make them easier to map onto the limited number of numeric keys. Within the system's internal pinyin parsing module, a set of medial vowel splitting rules is preset. When a medial vowel is input, the system first identifies its medial component and separates it from the final vowel. For example, for "ian", the system identifies "i" as the medial vowel and the remaining "an" as the final vowel. For the special medial vowels "uo" and "üe", this method chooses not to split them, treating them as a single final vowel.

[0184] The split pure vowels, along with the unsplit "uo" and "üe", are then merged into 11 sets of adaptable vowels (ɑ, ɑnɡ, engɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi). These vowel sets are optimized to balance pronunciation characteristics, usage frequency, and ease of key mapping. When defining the numerical vowel set, the system categorizes the split pure vowels and the unsplit "uo" and "üe" according to a preset "pronunciation tongue position consistency" principle. For example, "e", "üe", and "er" are grouped together because they have certain similarities in pronunciation or can share a key. "u", "o", "uo", and "i" are grouped together because this group of vowels may have commonalities in certain input scenarios or are used to achieve multi-phoneme triggering on specific keys. This merging strategy ensures that each digital key can cover a related set of vowels while maintaining the conciseness of the vowel set.

[0185] To provide flexibility and stability in key mapping for vowels, this method stipulates that the positions of ɑ and u / o / uo / i are fixed, en is only interchanged with e / üe / er, and the remaining vowels can be interchanged arbitrarily. These fixed and interchangeable rules are preset in the configuration interface or backend management module of the numerical vowel mapping table. For example, the vowel groups corresponding to the keys containing “ɑ” and “u / o / uo / i” cannot be changed. For the keys containing “en” and “e / üe / er”, the system may provide a limited number of interchange options. For other vowel groups, the system allows users to bind them to different numeric keys through dragging, selection, etc., and updates the “vowel-key identifier-encoding sequence” mapping table in real time.

[0186] Furthermore, when there is no need for adaptation and sharing, the positions of each vowel are unrestricted. When the user selects the "Custom Layout" mode, the system will remove the restrictions on the vowel positions, allowing the user to map any vowel group to any available numeric keypad. This mode is usually accompanied by a warning message informing the user that a custom layout may cause inconsistencies when synchronizing with the parent-child interface system or other devices. After the user confirms, the system will save their custom mapping table and apply it to the device.

[0187] Furthermore, this method also supports merging two groups of vowels into 10 compressed numerical vowel groups. A "Compressed Mode" option is provided in the system configuration. When the user selects this mode, the system will merge 11 groups of vowels into 10 groups according to preset merging rules (e.g., combining two vowels with relatively low usage frequency or similar pronunciations onto the same key). For example, "ɑi" and "ou" can be merged into the same key, or "ei / ü" and "ɑi" can be merged. After merging, the key will trigger the multiple vowels it contains by long-pressing or sliding.

[0188] Through the aforementioned technical solution, this method simplifies the types of finals by splitting medial vowels (excluding uo and üe) into medial vowels and finals, and merging them into 11 sets of adaptable finals. This allows a limited number of digital keys to cover a wider range of pinyin combinations. Specifically, fixing the positions of the high-frequency final ɑ related to medial vowels u / o / uo / i and restricting the transposition of en with e / üe / er ensures the stability and consistency of the core input experience, enabling users to quickly develop muscle memory. Simultaneously, allowing the remaining finals to be freely transposed, and with no restrictions on their positions when there is no need for adaptation sharing, greatly enhances the system's flexibility and user customization capabilities, allowing users to optimize the key layout according to personal preferences or specific scenarios. Furthermore, by providing the option to merge two sets of finals into 10 sets of compressed numerical finals, this method better adapts to small-screen devices such as smartwatches, providing a larger key area within limited screen space, effectively reducing accidental touches, and improving input efficiency and user experience on minimalist devices. This method maintains a concise and efficient set of vowels while also ensuring stability and high customizability, significantly improving the adaptability and user satisfaction of multilingual human-computer interaction.

[0189] In some of the above embodiments, a multilingual human-computer interaction method based on a number-ordered vowel interface is proposed. This method aims to provide an efficient and unified input method by defining a unique set of number-ordered vowels and mapping rules, and constructing an irreplaceable number-ordered vowel interface. However, in practical applications, users may have different preferences for the layout of the vowel keys, such as pursuing input efficiency, being accustomed to traditional keyboard layouts, or targeting specific learning scenarios. A single, fixed vowel sorting method is difficult to meet diverse user needs and may affect user experience and learning efficiency.

[0190] In this regard, this application further proposes that the ordering of the 11 sets of numerical vowels can be selected from any of the following schemes:

[0191] (1) Efficient input scheme: High-frequency vowels (ɑ, ɑn, en) are bound to keys 1-3, and low-frequency vowels (ei / ü, ou, ɑi) are bound to keys 9, 0 and 11;

[0192] (2) Alphabet keyboard compatibility scheme: Match and sort according to the order of the numeric keys on the alphabet keyboard;

[0193] (3) Children's teaching plan: Adjust the order of teaching and learning vowels according to pronunciation difficulty (simple vowels → compound vowels → nasal vowels).

[0194] This feature aims to provide flexibility and diversity in the arrangement of number-order vowels on the interface to adapt to the needs of different user groups and usage scenarios. By providing multiple preset sorting schemes, the system can automatically adjust the layout of the vowel keys according to the user's selection or specific scenarios, thereby optimizing input efficiency, improving user experience, or meeting teaching objectives. This selectivity ensures that the number-order vowel interface can better adapt to personalized and diverse application needs while maintaining the core encoding logic.

[0195] One efficient input method is a key layout strategy based on statistical principles and ergonomic design. Its core idea is to place frequently occurring vowels in Chinese Pinyin (such as a, an, en) on keys that are easier for users to reach (such as keys 1-3), while placing less frequently occurring vowels (such as ei / ü, ou, ai) on keys that are relatively farther away (such as keys 9, 0, 11). This layout aims to minimize the distance and time users spend moving their fingers during input, thereby significantly improving input speed and efficiency. For example, on mobile devices, users' thumbs or index fingers are usually more likely to reach keys in the center or bottom of the screen; binding high-frequency vowels to these areas can reduce cross-screen operations and improve the convenience of one-handed operation.

[0196] The alphabetic keyboard compatibility solution aims to provide a smooth transition for users accustomed to the traditional alphabetic keyboard layout. It matches the order of number vowels with the existing number key sequence on the alphabetic keyboard. For example, if the number 1 key on a traditional keyboard corresponds to a certain letter, the order of number vowels will also refer to this correspondence. This compatibility reduces the learning curve for users, allowing them to adapt to the new key layout more quickly when switching from a traditional input method to a number-vowel input method, reducing cognitive load and operational errors. For instance, if a user is accustomed to a specific arrangement of the number keys 1-0, the solution arranges the vowels according to their relative positions to these number keys on a traditional keyboard, preserving the user's existing muscle memory.

[0197] The children's learning program is specifically designed for children's language learning and teaching scenarios. It arranges the pronunciation of Pinyin vowels according to their difficulty, typically starting with the simplest simple vowels (such as a), gradually progressing to compound vowels (such as ao, ei) and nasal vowels (such as an, ang). This gradual progression aligns with children's cognitive development and learning curve, helping them systematically learn and master Pinyin pronunciation. By placing the less difficult vowels on easily recognizable and operable keys, the learning threshold for children is lowered, increasing their interest and efficiency. For example, in children's learning applications, the interface can first display simple vowels, gradually introducing compound vowels and nasal vowels after children have mastered them, with the key layout adjusted accordingly to enhance learning effectiveness.

[0198] By providing multiple selectable numerical vowel sorting schemes, this application effectively solves the problem that a single fixed vowel layout cannot meet the diverse needs of users. The efficient input scheme significantly improves input speed and efficiency by binding high-frequency vowels to easily accessible keys, reducing the user's operational burden. The letter keyboard compatibility scheme lowers the learning cost for users transitioning from traditional input methods to this method, allowing users to quickly adapt based on their existing muscle memory. The children's teaching scheme sorts vowels according to pronunciation difficulty, greatly optimizing children's experience in learning pinyin, helping them systematically master pronunciation rules, and improving teaching effectiveness. The introduction of these schemes allows the numerical vowel interface to more flexibly adapt to the personalized needs of different user groups while maintaining its core coding logic and cross-device compatibility, thereby comprehensively improving the efficiency, convenience, and user satisfaction of human-computer interaction.

[0199] In some embodiments described above, this application proposes a multilingual human-computer interaction method based on a number-sequential vowel interface. This method achieves unified encoding logic for multiple languages ​​by defining a unique set of number-sequential vowels and mapping rules, configuring cross-device compatible numeric keys and an extended key system, constructing an irreplaceable number-sequential vowel interface, and establishing a collaborative parent-child interface system. However, in practical applications, the screen sizes of different devices vary significantly. Using a single, fixed interface layout may result in overcrowded buttons and inconvenient operation on small-screen devices, while failing to fully utilize screen space to provide richer functionality on large-screen devices, thus affecting user experience and input efficiency.

[0200] In response, this application further proposes a dynamic keyboard that supports layered display. Specifically, this layered display mechanism divides the keyboard interface into a base layer, a standard layer, and an extended layer based on the device's screen size.

[0201] The dynamic keyboard's layered display capability refers to the virtual keyboard interface's ability to adaptively adjust its layout, content, and functionality based on the current device's screen size or type. The system dynamically loads and renders a keyboard layer that matches the screen size range by detecting the device's screen size (e.g., through an API provided by the operating system). This mechanism ensures that the user interface maintains optimal usability and functionality across different screen sizes, while the underlying number sequence and vowel mapping rules remain unchanged.

[0202] The base layer is designed specifically for small-screen devices (screen size 3.5 inches or less). In this layer, the keyboard interface retains only the core 1-12 keys, including extended keys 11 and 12. These extended keys are particularly important for small-screen devices because they enable single-key triggering of core vowels, thus providing efficient input within the limited screen space. To minimize obstruction of screen content, the base layer keyboard's screen height is strictly limited to 30% or less of the total screen height. This design ensures that users can still perform basic and accurate input operations on extremely small-screen devices such as smartwatches, while maintaining the visibility of the application interface.

[0203] The standard layer is suitable for mid-sized devices (screen sizes between 3.5 inches and 7 inches), such as mainstream smartphones. In this layer, the keyboard interface not only includes all core keys but also integrates tone keys and mode switching keys. The addition of tone keys allows users to directly input tones, improving input efficiency for languages ​​with tones. The mode switching keys facilitate quick switching between different input modes (such as abbreviated pinyin, full pinyin, and multilingual mode). The standard layer keyboard's screen-occupying height is set between 40% and 50% of the total screen height, aiming to provide a fully functional and comfortable input area while reserving sufficient display space for application content.

[0204] The extended layer is designed for large-screen devices (7 inches or larger), such as tablets or desktop touchscreens. This layer fully utilizes the larger screen space, offering richer advanced features, including shortcut bindings, reserved icons, and a single-key dual-phoneme candidate area. Shortcut bindings allow users to customize quick input methods for frequently used functions or phrases. Reserved icons can be used to display system status, input suggestions, or user-defined symbols. The single-key dual-phoneme candidate area visually displays dual-phoneme combinations triggered by a single key (e.g., a combination of a medial vowel and a tone), providing faster selection and input efficiency. The extended layer keyboard occupies 50% to 60% of the total screen height, designed to provide professional or high-efficiency users with an ultimate input experience and expanded functionality.

[0205] Through the above technical solution, the numerical vowel interface of this application can intelligently adjust its layout and functions according to the screen size of the device, thereby significantly improving user experience and input efficiency. On small-screen devices, the basic layer retains only the core keys and strictly controls the keyboard height to ensure usability and content visibility within a limited space, avoiding key congestion and accidental touches. On medium-screen devices, the standard layer provides comprehensive input functions, including tone keys and mode switching keys, meeting daily input needs while maintaining reasonable use of screen space. On large-screen devices, the extended layer further provides advanced functions such as shortcut key binding, reserved identifiers, and single-key dual-phoneme candidate areas, making full use of the larger screen space and greatly improving the input speed and convenience for professional and high-efficiency users. This layered display mechanism ensures the core consistency of the numerical vowel mapping rules across different devices while providing high interface adaptability, enabling users to obtain an optimized and efficient human-computer interaction experience regardless of the size of the device they use.

[0206] The aforementioned multilingual human-computer interaction method based on a number-sequential vowel interface, while defining a unique set of number-sequential vowels, a cross-device compatible key system, and a collaborative parent-child interface, and implementing unified multilingual encoding logic, still faces challenges in practical applications. These challenges include ensuring the consistency and continuity of the input experience, and optimizing the accuracy and user-friendliness of touch input on small-screen devices such as smartwatches. A lack of intelligent adaptation to external devices and refined optimization for small-screen touch input may lead to inconvenience for users switching between different input methods or devices, impacting overall input efficiency and satisfaction.

[0207] To address this, this application further proposes a method for optimizing the human-computer interaction experience. In this method, the dynamic keyboard and the external hardware keyboard are automatically detected via USB HID and Bluetooth HID protocols, with a detection latency of ≤100ms. Specifically, the system integrates USB HID and Bluetooth HID protocol stacks. When a new USB device is inserted or a Bluetooth device pairing request is detected, the system performs a handshake and device type identification through these protocols. Once identified as a hardware keyboard, the system immediately loads the corresponding driver or configuration and prepares to receive input, ensuring that the time from device connection to system recognition and readiness is extremely short.

[0208] Building upon this, the method supports automatic hiding / synchronous display / floating mode switching with a switching latency of ≤50ms. When the system detects that the hard keyboard is connected and active, the dynamic keyboard can automatically retract to free up screen space based on user preference or intelligent system judgment, or remain displayed and reflect the input status of the hard keyboard in real time. For example, when characters are entered on the hard keyboard, the corresponding keys on the dynamic keyboard will be highlighted synchronously to provide visual feedback to the user. In addition, the dynamic keyboard can also exist as a draggable and resizable floating window, allowing users to quickly click the soft keyboard for auxiliary input when needed, without obstructing screen content.

[0209] Furthermore, when a single-key dual-phoneme combination is input on the physical keyboard, the corresponding key on the dynamic keyboard is simultaneously highlighted. After receiving the key code from the physical keyboard, the system matches it against a preset single-key dual-phoneme combination mapping table. Once a match is successful, the system sends a command to the display module, causing the virtual key corresponding to that dual-phoneme combination on the dynamic keyboard interface to briefly highlight, thus intuitively informing the user of the currently input phoneme combination and enhancing the visual feedback and learning assistance when using the physical keyboard for complex input.

[0210] Considering the small screen size and high precision requirements of small-screen devices such as smartwatches, this application optimizes the touch control of the "11" and "12" extended buttons for these devices. This optimization supports adaptive touch area, with a minimum touch area of ​​≥8×8mm. The system intelligently expands the actual touch area to an effective recognition range of at least 8×8mm based on the user's touch point position and pressure distribution, ensuring correct recognition even with slight deviations in finger touch position, reducing accidental touches caused by large fingers or inaccurate operation. Simultaneously, this optimization includes an accidental touch filtering function, with a customizable pressure threshold ranging from 5-20g. The device screen integrates a pressure sensor that detects the force of the user's finger pressing the screen. Only when the pressure exceeds the user-defined threshold is it considered a valid touch operation, effectively distinguishing between intentional and unintentional touches. Through the above-mentioned optimization measures such as adaptive touch area and accidental touch filtering, the touch recognition accuracy can reach ≥98% under the test environment (room temperature 25℃±5℃, humidity 40%-60%, small screen device screen size ≤2 inches, processor ≥Snapdragon 855 / Apple A13, memory ≥2GB).

[0211] Through the above technical solutions, this application effectively solves the problems of inconsistent user experience and insufficient input accuracy on small-screen devices under multiple device and input modes. The automatic detection and mode switching function between the dynamic keyboard and the external hard keyboard allows the soft keyboard to respond intelligently when the hard keyboard is connected or disconnected, avoiding the tedious manual adjustments and ensuring a seamless input process. When the hard keyboard inputs single-key dual-phoneme combinations, the synchronous highlighting of the dynamic keyboard not only provides users with intuitive visual feedback, reducing the learning curve, but also enhances the certainty of the input process. Especially for small-screen devices such as smartwatches, the touch optimization of the "11" and "12" extended keys, including touch area adaptation and accidental touch filtering, significantly improves the touch recognition accuracy on limited screen space, effectively reduces accidental touches, and greatly improves the input experience on small-screen devices. This allows users to perform efficient and accurate number and vowel input even on small screens, thereby comprehensively improving the convenience and user satisfaction of multilingual human-computer interaction.

[0212] In some of the embodiments described above in this application, a multilingual human-computer interaction method based on a number-order vowel interface is proposed. This method constructs an irreplaceable number-order vowel interface by defining a unique set of number-order vowels and mapping rules, and configuring cross-device compatible digital key positions and extended key positions, thus realizing a unified encoding logic for multiple languages. However, when processing languages ​​with tones such as Chinese, relying solely on the combination of initials, medials, and vowels for input may lead to high ambiguity of candidate words, requiring users to spend more time selecting from a large number of homophones, thereby affecting input efficiency and accuracy.

[0213] In response, this application further proposes a number sequence tone expansion method, which includes a number sequence single tone function, a number sequence double tone function, and candidate prompt rules.

[0214] The single-tone input function allows users to directly input tone information using number keys after inputting the initial-medial-final combination. Specifically, numbers 1 to 4 correspond to the first to fourth tones in Mandarin, while number 5 corresponds to the neutral tone or no tone. Once the user presses the corresponding number key, the system simultaneously displays a list of candidate syllables with tones filtered by tone. For example, if the user inputs "han" and presses the number "1", the system will immediately display candidate words related to "hān"; if the user presses the number "3", it will display candidate words related to "hǎn". This direct tone input method allows users to quickly and clearly specify the tone of the desired syllable, thereby significantly reducing the length of the candidate list and improving the accuracy and efficiency of word selection.

[0215] The number-sequential dual-tone function adds a new set of number-sequential tone configurations to the number-sequential single-tone function. In this configuration, the number 6 corresponds to a neutral tone or no tone, while the numbers 7 to 0 correspond to the 4th, 1st, 2nd, and 3rd tones of Mandarin Chinese, respectively. Users can customize the order based on personal habits or high-frequency usage. By combining the number-sequential single-tone and number-sequential dual-tone functions to form a dual-tone system, the system can simultaneously suggest two sets of candidate syllables with tones. For example, after inputting a vowel-medial-vowel combination, the system can simultaneously display candidate lists filtered based on both the single-tone configuration (e.g., keys 1-5) and the dual-tone configuration (e.g., keys 6-0). This provides users with a wider range of choices, especially when users are uncertain about the tone of a syllable. They can explore two possible tone combinations simultaneously, thus finding the target word more flexibly and further improving the convenience of the input experience.

[0216] Furthermore, the candidate suggestion rules specify how the extended interface autonomously determines the maximum number of suggested strings per page based on various factors. These factors include ordinal vowels, key identifiers and extended key identifiers, ordinal strings, multilingual characters, and the device's screen size. For example, on smaller screen devices such as smartwatches, the extended interface intelligently reduces the number of candidate strings displayed per page and may increase character size to ensure clear readability; while on larger screen devices such as tablets or desktop monitors, more candidate strings can be displayed to increase information density. This dynamic adjustment mechanism ensures that the presentation of the candidate list is always optimal across different devices and usage scenarios, avoiding the user experience degradation caused by information overload on small screens while fully utilizing the display space on large screens, enabling users to efficiently browse and select candidate words.

[0217] Through the aforementioned technical solution, this application introduces a number-order tone extension function on top of the basic number-order vowel encoding system, effectively solving the problems of low input efficiency and candidate ambiguity caused by the lack of a clear tone input mechanism when inputting tones such as Chinese. The number-order single-tone function directly associates number keys with Chinese tones, enabling users to quickly and accurately input the tone information of syllables, thereby significantly narrowing the candidate range and improving input accuracy. Furthermore, the number-order dual-tone function provides additional tone configuration options and custom sorting capabilities, allowing the system to simultaneously prompt two sets of tone-containing candidate syllables. This provides great convenience and flexibility for users in scenarios where they are unsure of the specific tone or need to explore multiple pronunciation possibilities, effectively reducing the user's cognitive burden and trial-and-error costs. Simultaneously, the candidate prompting rules dynamically adjust the number of prompt strings per page based on factors such as device screen size and input content, ensuring that the candidate list always maintains optimal visual presentation and interactive experience across different devices and usage scenarios, avoiding information overload on small-screen devices or space waste on large-screen devices. These functions work together to enable the multilingual human-computer interaction method based on the number-sequence vowel interface to not only maintain the consistency and efficiency of the core encoding when processing languages ​​with tones, but also to achieve refinement and intelligence in tone input and candidate prompts, greatly improving the user input experience and overall input efficiency.

[0218] In some embodiments described above in this application, a numerical tone extension method is proposed, which realizes the synchronous display of a candidate list of tone-bearing syllables through numerical single-tone and numerical double-tone functions. However, in practical applications, different user groups (such as adults and children) have different frequencies of tone usage and cognitive habits. If a uniform and fixed tone key order is adopted, it may lead to low tone input efficiency for specific user groups, or even increase the input error rate, affecting the user experience.

[0219] In response, this application further proposes a custom sorting of the numerical dual tones, supporting users to set a "high-frequency tone priority" rule: in adult mode, the four tones of Mandarin (usage rate ≥35%) are bound to the 7 key, and in children's mode, the first tone and neutral tone (total usage rate ≥40%) are bound to the 6-7 key. The custom rule can be synchronized to the parent-child interface, and the tone input error rate is ≤3%.

[0220] Specifically, the customizable sorting of the numerical dual-tone function allows users to adjust the correspondence between number keys and Chinese tones (such as 1-4 tones and neutral tone) according to personal preferences or specific application scenarios. This customizable sorting mechanism breaks the fixed key bindings, providing flexibility to adapt to different input habits and efficiency needs. It supports users setting a "high-frequency tone priority" rule, allowing them to optimize the key layout based on the frequency of tone usage. For example, the most frequently used tones by users or specific groups can be bound to keys that are easier to access or remember, thereby improving input speed and accuracy. This can be achieved by statistically analyzing users' historical input data or preset linguistic statistics.

[0221] For adult users, considering the high frequency of use of the four tones in daily communication (e.g., usage rate of 35% or more), the system can prioritize binding the four tones to the 7 key in the numerical dual-tone system. This aims to optimize the input experience for adult users, reduce input paths for high-frequency tones, and improve input efficiency. For children, considering their language learning and usage characteristics, the combined usage frequency of the first tone and neutral tone in Mandarin may be high (e.g., combined usage rate of 40% or more). Therefore, the system can prioritize binding the first tone and neutral tone to the 6 and 7 keys in the numerical dual-tone system to adapt to children's cognitive habits and learning stages, reducing the difficulty and error rate of tone input. The custom tone sorting rules set by the user on any device can be seamlessly synchronized through the parent-child interface system. This means that regardless of whether the user is inputting on a mobile terminal, smart wearable device, or other industrial control equipment, their personalized tone key layout will remain consistent, ensuring the coherence of input logic across devices and the consistency of user experience. Through the above custom sorting and pattern optimization, this method aims to control the tone input error rate to 3% or lower. This indicates that the solution not only improves input efficiency but also effectively ensures input accuracy, thereby enhancing the overall quality of human-computer interaction.

[0222] By introducing a customizable dual-tone sorting function and supporting user-defined "high-frequency tone priority" rules, this method can personalize tone input based on the tone usage habits of different user groups. For example, for adult users, the high-frequency fourth tone is bound to easily accessible keys, while for children, the high-frequency first tone and neutral tone are bound to easily accessible keys, significantly improving the tone input efficiency and accuracy for specific user groups. Simultaneously, the synchronization of custom rules between parent and child interfaces ensures a consistent input experience across devices, effectively reducing tone input error rates, overcoming the limitations of fixed tone key layouts that cannot adapt to diverse user needs, and improving the overall convenience and user satisfaction of human-computer interaction.

[0223] In some of the embodiments described above in this application, candidate suggestion rules are proposed to autonomously determine the maximum number of suggested strings per page based on numerical vowels, key identifiers, extended key identifiers, numerical strings, and multilingual characters. However, in practical applications, the screen sizes of different devices vary greatly. If the display method of candidate suggestions is not optimized in a targeted manner, it may lead to information overload and difficulty in recognition on small-screen devices, while failing to make full use of screen space on large-screen devices, thereby affecting the efficiency and experience of user input.

[0224] To address this, this application further proposes a method for dynamically adjusting the number of string suggestions per page of the extended interface according to the screen size. Specifically, the extended interface refers to the area used to display candidate words, phrases, or syllables during user input. The dynamic adjustment of the number of string suggestions means that the system can intelligently calculate and set the maximum number of candidate strings that can be displayed per page based on the current device's screen physical size. This is typically achieved by obtaining screen resolution and physical size information through the operating system's application programming interface (API), and then calculating it in conjunction with a preset display strategy. For example, the system can maintain a mapping table between screen size and the number of recommended suggestions, or dynamically generate recommended values ​​based on factors such as screen area and pixel density using an algorithm.

[0225] This application provides specific adjustment schemes for devices with different screen sizes. For small-screen devices with a screen size of 2 inches or less, such as smartwatches, the system limits the number of strings displayed per page to 2 to 3. Simultaneously, to ensure readability on extremely small screens, the character size is set to no less than 16 points. To facilitate users browsing more candidates, these devices support switching candidate suggestion pages via vertical swipes. Vertical swipes are typically implemented through a touchscreen gesture recognition module; when the system detects a vertical swipe by the user in the candidate area, it loads and displays the previous or next page of the candidate list.

[0226] For smaller devices with screen sizes between 2 and 3.5 inches, such as some small smartphones or dedicated input devices, the system adjusts the number of strings displayed per page to 3 to 5. The character size is correspondingly adjusted to no less than 14 points to moderately increase information density while ensuring readability.

[0227] For mid-sized devices with screen sizes between 3.5 inches and 7 inches, such as mainstream smartphones or small tablets, the system sets the number of strings displayed per page to 6 to 8. The character size is further adjusted to no less than 12 points to make full use of the medium-sized screen space and provide a wider range of candidate options.

[0228] For large-screen devices with a screen size of 7 inches or larger, such as tablets or in-vehicle infotainment screens, the system expands the number of strings displayed per page to 9 to 11. The character size can be optimized based on the actual display effect, but is typically no smaller than size 12. To better suit the horizontal browsing habits of large screens, these devices support switching between candidate suggestion pages via left and right swipes. Left and right swipes are also implemented through a touchscreen gesture recognition module; when the system detects a horizontal swipe in the candidate area, it loads and displays the previous or next page of the candidate list.

[0229] It is important to emphasize that the candidate suggestion rules for all devices are strongly bound to the number-order vowel mapping rules and cannot be modified independently. This feature underscores the close relationship between the candidate suggestion display logic and the core number-order vowel encoding system. This means that regardless of screen size, the generation, sorting, and correspondence between candidate words and input codes must strictly adhere to the preset number-order vowel mapping rules. Users or third-party applications cannot modify or bypass this binding relationship to change the underlying logic of the candidate suggestions, thus ensuring the consistency, stability, and security of the input system across different devices and display modes. This strong binding is achieved through verification and restrictions in the system's core encoding module; any attempt to modify it will be blocked.

[0230] Through the above technical solution, this application can intelligently and dynamically adjust the number of string prompts and the character size of each page of the extended interface according to the screen size of different devices. This effectively solves the problem of incompatibility in candidate prompt display on screens of different sizes, avoiding information overload and difficulty in recognition on small-screen devices, as well as wasted space on large-screen devices. For example, on small-screen devices such as smartwatches, by limiting the number of prompts and increasing the character size, while supporting vertical swiping to switch pages, the user's reading comfort and ease of operation in limited screen space are greatly improved. On large-screen devices such as tablets, more candidates can be displayed by making full use of screen space, and the input efficiency is improved by swiping left and right to switch pages. In addition, the strong binding between the candidate prompt rules and the number sequence vowel mapping rules of all devices ensures the consistency of cross-device input logic and system stability, preventing the risk of confusion or tampering of the underlying coding logic due to display adaptation, thereby significantly optimizing the human-computer interaction experience in multi-device environments.

[0231] This application proposes a multilingual human-computer interaction method based on a number-order vowel interface. This method defines a unique set of number-order vowels and mapping rules, configures cross-device compatible numeric keys and an extended key system, constructs an irreplaceable number-order vowel interface, and a collaborative parent-child interface system, achieving a unified encoding logic for multiple languages ​​and emphasizing the indivisibility of the core technology. However, in practical applications, especially for languages ​​with tones such as Chinese, efficiently and flexibly inputting tone information while maintaining compatibility with the core number-order vowel encoding system is crucial for improving user experience and input efficiency. Existing solutions may have limitations in tone input methods, such as requiring mode switching, occupying too many keys, or being complex to operate, which increases the user's cognitive burden and input interruption.

[0232] In response, this application further proposes a dual-mode tone interface human-computer interaction method, the specific implementation of which includes:

[0233] First, configure the character tone settings. This configuration includes two types: dedicated tone keys and shared keys. Dedicated keys are keys specifically for inputting tones, and their size can be designed to match that of regular numeric keys to ensure user continuity and familiarity. These dedicated keys can be flexibly adapted to various layouts based on device screen size and user habits, such as a 4x4 matrix layout or a more compact circular layout, to optimize user access efficiency. For example, on touchscreen devices, they can be designed as small, independent areas that pop up above or to the side of the main keyboard area. Shared keys allow tone functions to share the same physical or virtual keys with other characters such as punctuation marks, effectively saving interface space, especially suitable for small-screen devices such as smartwatches. For example, the shared key mapping relationship can be set to ",-T1, .-T2, ?-T3, !-T4" and / or ",-T1, .-T2, / -T3, ]-T4 or \-T4", which means that specific punctuation mark keys can output the corresponding tone mark under specific operations (such as long press or slide) (T1 represents the first tone, T2 represents the second tone, and so on).

[0234] Secondly, seamless expansion of the consonant set is achieved. This method supports two input rules without switching input modes. The first rule is "initial consonant + medial vowel + tone + final vowel," allowing users to input in the order of initial consonant, medial vowel, tone, and final vowel. For example, inputting "b" + "i" + "T1" + "a" represents the first tone of "bia." The second rule is "initial consonant + tone + medial vowel + final vowel," allowing users to input the tone immediately after the initial consonant, followed by the medial vowel and final vowel. For example, inputting "b" + "T1" + "i" + "a" represents the first tone of "bia." These two rules share the numerical final vowel mapping and candidate logic with the tonic consonant set, ensuring that the system still processes based on the numerical final vowel encoding system during consonant set input and uses a shared candidate word library to generate a candidate list, thereby guaranteeing the consistency of input logic and the accuracy of candidate results.

[0235] Furthermore, it achieves dual-mode collaboration. This method allows for free switching between numerical and character tones to adapt to different user preferences and usage scenarios. Both tone modes share the same tone-matching vocabulary mapping library. This means that regardless of whether the user inputs the same tone identifier using numerical tones (e.g., inputting a tone via number keys) or character tones, the system can retrieve and output consistent candidate results from the same vocabulary library, ensuring consistency of candidate words across different input modes. When a conflict arises between the two modes in recognizing tone input, the system prioritizes the operation that was triggered first, ensuring the clarity of the input behavior. In addition, the system has an automatic memory function, automatically remembering the tone mode the user has recently used. For example, if the user last used the character tone mode, the system will automatically switch to that mode the next time the input method is launched, thereby improving the convenience of the user experience.

[0236] Through the above technical solutions, this application further enhances the flexibility and efficiency of tone input based on the numerical vowel core encoding system proposed in claim 1. The introduction of character tone configuration, especially the flexible design of independent and shared keys, allows users to choose the most suitable tone input method according to the device screen size and personal habits, effectively solving the problem of limited tone input space on small-screen devices and avoiding the hassle of frequently switching input modes due to tone input. Seamless expansion of the consonant set, by supporting two initial-medial-vowel combination orders, greatly improves the adaptability of input rules, giving users more choices when inputting syllables with tones, thereby improving input speed and accuracy. The dual-mode collaborative mechanism ensures seamless integration of numerical tone and character tone input methods, the shared vocabulary mapping library guarantees the consistency of candidate results, and the conflict handling and pattern memory functions further optimize the user experience, enabling users to complete tone input in the most natural and efficient way in different scenarios, significantly reducing learning costs and operational complexity.

[0237] In some embodiments described above, a dual-mode tone interface human-computer interaction method is proposed, which configures character tones, including independent tone keys and shared keys. However, when using shared keys for "punctuation-tone," effectively distinguishing whether the user intends to input punctuation marks or tones within the limited key space, and ensuring the intuitiveness, efficiency, and personalized user experience of the input operation, is a technical problem that needs to be solved. A lack of a clear distinction mechanism may lead to user confusion during input, reducing input efficiency and accuracy.

[0238] In response, this application further proposes the following triggering logic for the shared "punctuation-tone" key: a light touch (press duration ≤ 200ms) outputs a punctuation mark, a long press (press duration ≥ 300ms, ≥ 400ms in child mode) pops up a tone candidate list (T1-T4), and the tone is output by sliding to the corresponding tone key and releasing it during the long press; the application also supports user-defined light touch / long press duration thresholds (range 100-500ms).

[0239] Specifically, this triggering logic for the shared punctuation-tone key aims to resolve the ambiguity when a single physical key carries two different functions (punctuation input and tone input). By distinguishing the user's press duration and subsequent actions, the system can accurately determine the user's input intent, thus enabling convenient input of both functions without adding extra keys or switching input modes. When the user presses the shared key for a short time, i.e., a press duration of less than or equal to 200 milliseconds, the system recognizes it as a punctuation input operation. This design aligns with users' daily habit of quickly and frequently inputting punctuation marks, ensuring the immediacy and fluency of punctuation input. When the user presses the shared key for a longer time, i.e., a press duration of greater than or equal to 300 milliseconds, the system recognizes it as a tone input intent. At this time, a candidate list containing tone options such as T1-T4 will pop up on the interface. Specifically, for the children's mode, the long-press threshold is set to greater than or equal to 400 milliseconds. This is to account for the possibility that children may have inconsistent press durations during operation. Appropriately extending the threshold can reduce accidental touches and improve input accuracy in children's mode. After the tone candidate list pops up, users do not need to lift their fingers. They only need to slide their fingers to the corresponding tone key in the candidate list while maintaining the long press, and then release their fingers to complete the tone input. This sliding selection method makes the tone input process more coherent and intuitive, avoiding multiple clicks or finger lifts, and improving input efficiency. To accommodate different users' usage habits and operating preferences, the system allows users to customize the press duration threshold for light touches and long presses within the range of 100 milliseconds to 500 milliseconds according to their own needs. This feature greatly enhances the personalization and user-friendliness of the system, allowing users to adjust the trigger sensitivity according to their own operating habits and further optimize the input experience.

[0240] By employing the triggering logic of the shared "punctuation-tone" key, this application effectively resolves the ambiguity issue arising from the reuse of punctuation and tone input functions under limited key positions. Users can clearly distinguish their input intent through intuitive tap and long press operations, avoiding input errors or reduced efficiency caused by functional confusion. Tap operations enable rapid input of punctuation marks, aligning with users' daily input habits; while the combination of long press and swipe selection makes tone input more coherent and intuitive, allowing convenient selection of the desired tone without switching input modes, greatly enriching the functionality of a single key. Furthermore, optimization of the press duration threshold for child mode and the ability to support user-defined tap / long press duration thresholds further enhance the system's adaptability and personalization, ensuring a smooth and accurate input experience for users of different ages and operating habits. This triggering logic, working in conjunction with the dual-mode tone function, ensures that tone input can be conveniently and efficiently integrated into the overall encoding logic in multilingual human-computer interaction methods, thereby improving the overall usability and intelligence of the system.

[0241] In some embodiments described above in this application, a seamless consonant set expansion function is proposed, allowing users to input combinations of initials, medials, character tones, and finals using different input rules without switching input modes. However, in practical applications, how to intelligently determine user intent and automatically activate consonant set expansion to improve input efficiency and user experience, and avoid the tediousness of manual operation or mode switching, remains a problem that needs to be solved. To address this, this application further proposes the activation method for the aforementioned consonant set expansion, specifically including the following scenarios:

[0242] First, if no further input is received within 500ms after inputting "initial consonant + medial vowel," the system will automatically display a consonant candidate list. This mechanism aims to intelligently determine the user's input intent. After the user completes the input of the initial consonant and medial vowel combination, the system's internal input monitoring module starts a timer. If no subsequent input event (such as a vowel, tone, or space) is detected within the set 500ms time, the system will determine that the user may need candidate words from the consonant candidate list, thus automatically triggering the consonant candidate list pop-up logic. This logic will retrieve and display matching candidate words from a preset consonant candidate list based on the currently input initial consonant and medial vowel, avoiding the need for the user to manually trigger the consonant list and thus improving input efficiency.

[0243] Secondly, in multilingual input scenarios, the system will automatically parse according to consonant set rules. This means that when users input Chinese and other languages ​​in combination, the system can intelligently identify and apply consonant set rules without requiring users to manually switch language modes or input methods. The system's built-in language recognition module can determine in real time whether the current input belongs to Chinese Pinyin or another language based on information such as the character sequence, word frequency statistics, and contextual semantics. When it detects that the user is performing multilingual input, such as inputting English letters followed immediately by Pinyin initials, or inserting English words during Pinyin input, the system will automatically activate consonant set rule parsing, attempting to match and parse according to the rules of the consonant set to provide more accurate candidate words or phrases.

[0244] Furthermore, in children's teaching scenarios, the consonant candidate set will simultaneously display an animated breakdown of the pinyin into "initial consonant + medial vowel + tone." This is a special optimization for children's teaching scenarios, helping children understand the composition of pinyin through visualization. In children's teaching mode, when the consonant candidate set is triggered and displayed, the display module will simultaneously call an animation rendering module. This animation module will break down the pinyin syllables in the current consonant candidate set into three parts: initial consonant, medial vowel, and tone, and display them dynamically and visually (e.g., different colors highlighting, sequential appearance, or animations with pronunciation lip shape prompts) near the candidate area or in a separate teaching area. This visual breakdown helps children intuitively understand the composition structure of pinyin, enhancing their learning interest and effectiveness.

[0245] Through the above technical solutions, this application achieves intelligent and automated activation of consonant set expansion, significantly improving user input efficiency and experience in different scenarios. Specifically, after the user inputs an initial consonant and a medial vowel, the system can intelligently determine the user's intent and automatically pop up consonant set candidates if there is no subsequent input within 500ms, avoiding the tedious manual triggering by the user and making the input process smoother. In multilingual mixed input scenarios, the system can automatically parse the input according to the consonant set rules without requiring the user to manually switch language modes, greatly simplifying the complexity of cross-language input and improving the accuracy and convenience of mixed input. In addition, for children's teaching scenarios, the consonant set candidates simultaneously display the pinyin splitting animation of "initial consonant + medial vowel + tone", effectively helping children understand the composition of pinyin through an intuitive visual teaching method, improving learning interest and effectiveness. These intelligent activation and adaptation mechanisms make the consonant set expansion function more in line with user needs and further optimize the intelligence level of multilingual human-computer interaction.

[0246] In some of the embodiments described above in this application, a multilingual human-computer interaction method based on a number-sequential vowel interface is proposed. This method achieves efficient and unified input by defining a unique set of number-sequential vowels and mapping rules, configuring cross-device compatible numeric keys and an extended key system, constructing an irreplaceable number-sequential vowel interface, and establishing a collaborative parent-child interface system. However, in practical applications, users have higher demands for the flexibility and diversity of input methods, especially under different devices and usage scenarios. A single-key input mode may not fully meet users' personalized interaction habits and convenience requirements.

[0247] In response, this application further proposes a generalized, dynamically expandable interface human-computer interaction method, including:

[0248] 1) Flexible wake-up methods: Supports shortcut keys, voice commands, gestures, floating icons, or user-defined trigger methods. On Android 8.0 and iOS 12.0 and above mobile terminals, the trigger response latency is ≤80ms (test environment: room temperature 25℃, no external interference); the voice wake-up word can be customized, and the recognition accuracy is ≥92% in ambient noise below 50dB and in the frequency range of 200-3000Hz. Offline wake-up is supported.

[0249] 2) Core interface configuration: Includes associated prompt area, dynamic candidate area, fault-tolerant prompt area, and multimodal interaction area. The multimodal interaction area is compatible with the input logic mapping of hard keyboard, static soft keyboard, and dynamic soft keyboard.

[0250] 3) Multimodal Interaction Extension: Supports multimodal input and output such as speech-to-text, text-to-speech, gesture control, and image character recognition. Multimodal data needs to be mapped to the corresponding numerical vowel codes, which are then used in conjunction with key input codes to generate candidate data and synchronized to the dynamic candidate area in real time. Among them, speech-to-text supports ≥8 commonly used languages, with an online recognition accuracy of ≥95% and an offline recognition accuracy of ≥90%. Image character recognition supports multilingual characters with an accuracy of ≥90%.

[0251] 4) Adaptation and optimization: The interface can be dragged and scaled (range 50%-200%), the font size of the dynamic candidate area is adjustable from 8 to 24 points, and it supports switching between dark / light / high contrast modes; the dynamic candidate area and the trigger operation area are horizontally aligned, with an alignment error of ≤1 pixel.

[0252] The aforementioned flexible activation methods refer to the various non-traditional button methods users can use to launch or access the dynamically extended interface, adapting to different scenarios and user preferences to improve operational convenience. Specifically, shortcut keys can be combinations of physical buttons on the device or specific key combinations on the soft keyboard. When a user presses a preset shortcut key, the system listens and triggers the interface display. Voice commands allow users to speak preset wake words, which the system recognizes using a built-in voice recognition module. This module maintains high recognition accuracy in low-noise environments, supports offline recognition, and the wake word can be customized by the user. Gestures refer to specific gestures performed by the user on the screen. The touchscreen driver captures the gesture data, which is then parsed by the gesture recognition module. When a preset gesture is recognized, the interface is triggered. Floating icons display a clickable icon on the edge of the screen or in a specific area; clicking the icon activates the interface. Furthermore, the system supports user-defined trigger methods, such as automatic activation when shaking the device or connecting specific peripherals. In mainstream mobile operating systems, the response latency from trigger operation to interface display is strictly controlled within 80ms to ensure a smooth user experience.

[0253] The core interface configuration described above refers to several key functional areas within the dynamically expandable interface. These areas work together to provide users with comprehensive input assistance and interactive feedback. The contextual suggestion area displays information related to the current input context or user habits, such as frequently used phrases, suggested words, or the next character / vocabulary predicted based on the user's input history. The dynamic candidate area is the core output area, displaying a real-time list of candidate characters, words, or phrases generated based on user input (whether keystrokes, voice, or other modalities). Its content, layout, and size can be dynamically adjusted according to the input state and screen space. The error-tolerant suggestion area provides corrective suggestions or prompts when the user makes an input error; for example, when the system detects that the user may have pressed the wrong key, this area displays the most likely correct input option. The multimodal interaction area integrates multiple input methods. It is compatible with the input logic mapping of traditional hard keyboards, static soft keyboards, and dynamic soft keyboards, and also provides entry points and feedback for non-keystroke inputs such as voice, gestures, and image recognition.

[0254] The aforementioned multimodal interaction extension refers to the interface's ability to not only support traditional key input but also process and integrate various non-key input / output modes, greatly expanding the dimensions and convenience of human-computer interaction. The speech-to-text function allows users to input speech via microphone, and the system converts the speech signal into text in real time. This function supports multiple commonly used languages ​​and has high-accuracy online and offline recognition capabilities. The converted text is mapped to the corresponding numerical vowel codes and sent to the candidate generation module along with the key input codes. The text-to-speech function can read aloud the text in the candidate list or the text entered by the user, providing convenience for visually impaired users or users requiring auditory feedback. Gesture control, in addition to waking up the interface, can also be used for operations within the interface, such as swiping to select candidate words, pinching to zoom the interface, etc. Gesture data is also parsed and mapped to corresponding operation commands or codes. The image character recognition function allows users to capture images containing text using the device's camera. The system uses OCR technology to recognize the characters in the image, converting the recognized characters into numerical vowel codes, which then participate in candidate generation. All modal input data will be uniformly mapped to number-order vowel codes, then processed collaboratively to generate a unified candidate list, which will be displayed in real time in the dynamic candidate area.

[0255] The aforementioned adaptation and optimization refer to the interface's high degree of adaptability and customizability, capable of adjusting to different devices, user habits, and environments to provide the best visual and operational experience. The interface supports dragging and scaling, allowing users to freely move the interface to avoid obscuring important content on the screen. Simultaneously, the interface size can be scaled from 50% to 200% to accommodate different users' visual needs or operating habits. The font size in the dynamic candidate area can be adjusted between 8 and 24 points, ensuring clear display across different screen sizes and user preferences. The interface also supports switching between dark / light / high-contrast modes, such as switching to dark mode at night or in low-light environments to reduce eye strain. The dynamic candidate area intelligently aligns horizontally with the user's trigger area, ensuring the candidate list appears near the user's visual focus, reducing eye movement, improving input efficiency, and controlling alignment errors to within 1 pixel.

[0256] Through the aforementioned technical solutions, users can conveniently access the input interface via shortcut keys, voice commands, gestures, or floating icons, depending on different usage scenarios and personal habits, greatly improving operational convenience and user experience. The core interface configuration, including the associated prompt area, dynamic candidate area, error-tolerant prompt area, and multimodal interaction area, provides users with comprehensive and intelligent input assistance and feedback. Especially in terms of multimodal interaction expansion, it supports multiple input / output methods such as speech-to-text, text-to-speech, gesture control, and image character recognition, allowing users to move beyond traditional key input and interact with the computer in a more natural and efficient manner. All this multimodal data can be seamlessly mapped to numerical vowel codes, co-generating candidates with key input to ensure the consistency of input logic. Furthermore, through adaptation optimizations such as draggable and zoomable interfaces, adjustable font sizes, and multiple mode switching, the interface can adapt to different device screen sizes and user preferences, ensuring precise alignment between the dynamic candidate area and the trigger operation area, thereby significantly improving input efficiency and visual comfort. This makes the numerical vowel code-based input method achieve a higher level of versatility and user-friendliness.

[0257] In some of the embodiments described above in this application, a multilingual human-computer interaction method based on a number-sequential vowel interface is proposed, and a generalized dynamically extended interface is further introduced, including an associated prompt area, a dynamic candidate area, a fault-tolerant prompt area, and a multimodal interaction area to support multiple input methods. However, in practical applications, whether input is made via keypad or multiple modalities such as voice and gestures, mis-touch or mis-recognition may occur, which will affect the accuracy and efficiency of user input and reduce user experience.

[0258] In response, this application further proposes that the error-tolerant prompt area calculates the probability of mis-touch / mis-recognition based on the physical distance of the button position or the similarity of voice recognition, and the accuracy of the Top 1 correction candidate is ≥92%, and the correction logic is consistent with the number sequence vowel mapping.

[0259] The fault tolerance prompt area, as a part of the generalized dynamic extended interface, its core function is to intelligently identify and process potential errors in user input. Specifically, when the user makes a key input, the system can calculate the probability of the user accidentally touching an adjacent key based on the physical distance between the actual touch point of the user and the center point of the preset key, combined with the key size, the user's finger model, and historical input data. For example, if the user intends to input the key "1", but the touch point is closer to the key "2", the system can evaluate the possibility of accidentally touching "2" based on this. When the user makes a voice input, the system calculates the misrecognition probability of each candidate result according to the similarity between the multiple candidate results output by the speech recognition engine and the user's actual pronunciation, as well as the matching degree with the number sequence vowel encoding rules. For example, if the user pronounces "an", but the speech recognition result contains "ang" and the acoustic similarity is high, the system can calculate the probability that "ang" is a misrecognition. Through these two methods, the system can quantify the degree of input errors and provide data support for subsequent corrections.

[0260] After calculating the probability of accidental touch or misrecognition, the system will generate a correction candidate list and select the one with the highest probability or the most in line with the user's intention as the Top1 correction candidate. This application requires the accuracy rate of this Top1 correction candidate to reach or exceed 92%. To achieve this goal, when selecting the Top1 correction candidate, the system not only considers the probability calculation results, but also comprehensively uses various information such as context analysis, user historical input habits, word frequency statistics, and language models for intelligent judgment and sorting. For example, when the user inputs "ping an", even if the initial recognition deviates, the system can prioritize recommending "ping an" as the correction result based on high-frequency words and context, thus ensuring the accuracy of the correction.

[0261] The correction logic of this application is highly consistent with the number sequence vowel mapping, which means that all error correction operations must strictly follow the "vowel - key position identifier - coding sequence" trinity mapping table defined by this method. When the system determines that there is an accidental touch or misrecognition, it will not perform a simple character replacement outside the coding framework of the number sequence vowel, but will infer the possible true input intention of the user based on the number sequence vowel mapped by the key position, its coding sequence, and the voice feature mapping relationship. For example, if a key position is determined to be accidentally touched, the system will first consider the number sequence vowels corresponding to this key position and its adjacent key positions, and combine its coding sequence for correction. For speech recognition errors, the correction will also preferentially match those vowel combinations that exist in the number sequence vowel set and have a high acoustic similarity. This consistency ensures that the correction result is not only correct on the surface, but also completely compatible and effective in the underlying coding logic, thus maintaining the stability of the entire multilingual human - machine interaction method and the integrity of the core coding system.

[0262] By introducing a mechanism that calculates the probability of accidental touches / misrecognitions based on physical distance between keys or speech recognition similarity, this application can intelligently identify potential errors in user input and quantify their likelihood. Combined with a Top-1 correction candidate accuracy rate of ≥92%, the system can provide highly reliable preferred correction suggestions, significantly reducing the frequency of manual corrections due to input errors, thereby improving input efficiency and user experience. More importantly, the correction logic is consistent with the number sequence vowel mapping, ensuring seamless integration of the error correction process with the core encoding system of this method, avoiding conflicts between the correction results and the underlying encoding logic, and maintaining the stability and consistency of the entire multilingual human-computer interaction method. This allows users to receive timely and accurate corrections even for minor accidental touches or speech recognition deviations when using the number sequence vowel interface for multimodal input, greatly enhancing the system's robustness and ease of use, especially on small-screen devices or in noisy environments, effectively improving input accuracy.

[0263] In some of the above embodiments, a general dynamic extended interface human-computer interaction method is proposed, which includes a multimodal interaction area, is compatible with multiple input and output methods, and maps multimodal data to corresponding numerical vowel codes. However, in practical applications, different user groups (such as adults and children) or specific input scenarios (such as voice input with syllable tones) have different requirements for the recognition accuracy, error tolerance, and feedback mechanism of multimodal interaction. If a single interaction mode is used, it may lead to low recognition efficiency or poor user experience, and the advantages of multimodal interaction cannot be fully utilized.

[0264] In response, this application further proposes a mode adaptation scheme for the multimodal interaction area. This scheme includes an adult mode, a child mode, and a single-key dual-phoneme collaborative function.

[0265] The adult mode is designed to provide optimized services for professional users or those inputting information in specific domains. In this mode, the system enhances the recognition of professional terms input by integrating industry-specific thesaurus, domain knowledge graphs, and contextual semantic analysis models. For example, in the medical field, the system can prioritize recognizing and completing medical terminology; in the financial field, it can accurately identify financial terms and abbreviations. This enhanced recognition mechanism ensures that in complex professional contexts, the multimodal interaction area can recognize and process professional terms with higher accuracy (e.g., 90% or higher), thereby significantly improving the input efficiency and accuracy for professional users.

[0266] The Kids Mode is designed specifically for children, taking into account their unique pronunciation characteristics (such as non-standard pronunciation and uneven speech speed) and learning needs. In this mode, the system optimizes the speech recognition module by employing broader acoustic and language models, enhancing its tolerance for non-standard pronunciation. Even with pronunciation errors, the system can still achieve high accuracy (e.g., 88% or higher). Furthermore, for scenarios such as children reading picture books, the system integrates optimized image character recognition, efficiently and accurately recognizing text in picture books (e.g., 90% or higher). It provides good recognition results even with mixed text and image layouts or diverse fonts, assisting children in learning and interaction.

[0267] The single-key dual-phoneme collaboration feature aims to enhance the integration of voice and keyboard input. When a user inputs a syllable with a tone via voice, the system immediately recognizes the syllable and its tone. Subsequently, the key corresponding to that syllable on the dynamic keyboard interface (especially keys involving single-key dual-phoneme combinations, such as the combination of the medial vowel i or u / ü with a tone, i.e., tone marking medial vowels) is simultaneously highlighted. This visual feedback mechanism not only intuitively confirms the result of the voice input to the user but also guides the user to understand the mapping relationship between voice and keys. Especially when learning or using single-key dual-phoneme input rules, it effectively reduces the learning cost and improves input efficiency and accuracy.

[0268] Through the aforementioned technical solutions, this application provides a refined mode adaptation function for the multimodal interaction area, effectively solving the technical problems of insufficient recognition accuracy, lack of error tolerance, and unintuitive interactive feedback under different user groups and specific input scenarios. Specifically, the adult mode significantly improves the input efficiency and accuracy of professional users in complex contexts by strengthening the recognition of professional terms; the children's mode provides a more user-friendly and efficient learning and interactive experience for children by enhancing pronunciation error tolerance and optimizing picture book character recognition; and the single-key dual-phoneme collaborative function, through the visual synchronous feedback of voice recognition and dynamic keyboard, intuitively guides users to understand and master complex phoneme combination input, thereby reducing the learning threshold and improving overall input efficiency and user satisfaction. These mode adaptation functions enable the multimodal interaction area to provide a more intelligent, personalized, and efficient human-computer interaction experience according to specific usage scenarios and user characteristics.

[0269] This application proposes a multilingual human-computer interaction method based on a number-order vowel interface. It lays a foundation for efficient and unified input by defining a unique set of number-order vowels and mapping rules, configuring cross-device compatible numeric keys, constructing an irreplaceable number-order vowel interface, building a collaborative parent-child interface system, and implementing unified multilingual encoding logic. However, in practical applications, facing a wide variety of smart devices, diverse usage scenarios, and increasing multilingual demands, relying solely on a single universal interface and interaction logic may be insufficient to fully meet the input habits of different device types, operational efficiency in specific scenarios, and deep support for a broader range of languages, thus affecting user experience and the system's universality.

[0270] To address this, this application further proposes a human-computer interaction method that integrates scene adaptation, multi-directional layout adaptation, and core engine support. This method first achieves cross-device adaptation, specifically by configuring the effective touch area of ​​the keypad differently based on device type. For example, for mobile terminals, the effective touch area is set to no less than 8×8mm to accommodate the typical touch precision of fingers. For smart wearable devices, including smartwatches, the effective touch area is configured to no less than 10×10mm, and extended keys 11 and 12 also follow this standard to compensate for the inconvenience of operation caused by small screens. Commercial devices and industrial control panels are configured with touch areas of no less than 12×12mm and no less than 15×15mm, respectively, to ensure accuracy and reliability in more complex or harsh environments. Throughout this process, the core input logic and number sequence vowel mapping rules remain completely consistent across all devices. Only the physical size and layout density of the keys are allowed to be adjusted, while the core coding relationships cannot be modified, thus ensuring the uniformity of the cross-device input experience and the stability of the coding logic.

[0271] Secondly, this method supports multi-directional layout adaptation, providing various interface layout options to suit different device postures and user preferences. Specifically, it supports right-angled left-aligned vertical layout, right-angled right-aligned vertical layout, and top-aligned reverse layout. In the top-aligned reverse layout, the character direction can be manually selected by the user or automatically adapted according to the current document's layout direction. This flexible layout switching significantly improves user visual comfort and operational convenience in different scenarios without affecting input efficiency, and the switching latency is controlled within 50ms, ensuring a smooth user experience.

[0272] Furthermore, this method achieves scenario-specific adaptation, with functional optimizations and customizations tailored to specific application scenarios. In children's educational scenarios, the system features animated pronunciation and lip-syncing displays to help children learn pronunciation intuitively; it also offers three adjustable speech speeds to suit different learning stages and includes follow-up feedback for real-time pronunciation correction. In industrial control scenarios, the system optimizes the input flow and caching mechanism of core commands to improve operational efficiency and response speed. In cross-border communication scenarios, it enhances multilingual offline translation and pronunciation prompts based on numerical vowel encoding, effectively overcoming barriers to cross-language communication.

[0273] Furthermore, this method possesses multilingual extension capabilities, aiming to extend it to more languages. This is achieved by binding core characters of other languages ​​to reserved encoding positions for ordinal vowels based on phonetic features. In addition, the system supports providing standard pronunciation prompts for multilingual words, with a naturalness score of no less than 4.0, ensuring the accuracy and fluency of pronunciation, thereby greatly expanding the application scope of this method in multilingual environments.

[0274] Finally, this method integrates a core support engine, providing underlying support for the entire interactive approach. It includes an offline encoding verification module based on number-order vowel encoding to verify the validity of the input encoding; a mixed-mode parsing module capable of handling data under different input modes; an offline lexicon indexing module providing fast and accurate vocabulary retrieval services; and a multilingual bidirectional offline translation function, supporting translation between multiple languages ​​in offline environments. The engine's built-in offline lexicon is massive, containing no fewer than 80,000 entries for Chinese, no fewer than 50,000 for English, and no fewer than 30,000 for each of other languages, ensuring the stability and efficiency of all functions in offline mode.

[0275] Through the aforementioned cross-device adaptation, multi-directional layout adaptation, scenario-customized adaptation, multi-language extension adaptation, and core support engine settings, this method effectively addresses the limitations of general input solutions in terms of diverse devices, complex usage scenarios, and multi-language support. Specifically, the differentiated keypad touch areas ensure a comfortable and accurate input experience for users on various devices, from small smartwatches to large industrial control panels, while maintaining the consistency of core coding logic. Multi-directional layout adaptation provides flexible interface display methods, significantly improving user visual comfort and operational efficiency in different operating environments. Customized functions for specific scenarios such as children's education, industrial control, and cross-border communication greatly enhance the system's practicality and professionalism, enabling it to accurately meet the needs of specific users. Multi-language extension adaptation and standard pronunciation prompts allow this method to seamlessly support more languages, broadening its global application scope. Finally, the core support engine's offline encoding verification, mixed-mode parsing, offline dictionary indexing, and multi-language bidirectional offline translation functions ensure efficient and stable operation of the system in offline environments, greatly improving system robustness and user experience. These synergistic technical solutions have enabled this approach to evolve from a basic input framework into a highly adaptable, feature-rich, and stable multilingual human-computer interaction solution.

[0276] In some of the embodiments described above in this application, a multilingual human-computer interaction method based on a number-sequence vowel interface is proposed and customized for children's teaching scenarios. However, in the actual process of children's language learning, simply configuring scene functions may not be sufficient to provide sufficiently intuitive and interactive teaching assistance, especially in pronunciation learning and input practice. Children may find it difficult to accurately grasp the pronunciation essentials or maintain their learning interest, thereby affecting the learning effect.

[0277] In response, this application further proposes a visual teaching function for children's teaching scenarios, which includes animated display of the pronunciation mouth shape of Chinese Pinyin, three-level speech speed adjustment (50 / 100 / 150 words / minute), interactive syllable reading games, encouraging feedback after correct input, and single-key dual-phoneme combination support for visual split display.

[0278] Specifically, the animated demonstration of mouth shapes for Pinyin pronunciation aims to visually and intuitively present the changes in key elements such as mouth shape and tongue position during Pinyin pronunciation. For example, the system can preset or generate 2D or 3D animations in real time to simulate the mouth shape changes of standard speakers, helping children understand and imitate correct pronunciation postures. For beginners, this can compensate for the shortcomings of purely auditory learning, providing multi-sensory input and thus improving pronunciation accuracy. Three adjustable speech speeds allow users to flexibly adjust the speech speed of the pronunciation demonstration according to the learner's age, cognitive level, and learning progress. For example, beginners can choose a slow speed of 50 words per minute so they can clearly hear and distinguish each syllable; as learning progresses, this can be gradually increased to 100 words per minute or 150 words per minute to adapt to a more natural speech speed. This personalized speech speed setting helps children gradually master pronunciation rhythm and fluency. The interactive syllable repetition game encourages children to practice pronunciation in a gamified way. For example, the system can play standard syllables and guide children to repeat them, using a built-in speech recognition module to analyze and score their pronunciation in real time. Games can be designed as level-based or scoring challenges to increase the fun and interactivity of learning, stimulating children's enthusiasm for actively participating in pronunciation practice. Providing encouraging feedback after correct input aims to offer children immediate and positive feedback to reinforce correct learning behaviors. For example, when a child successfully pronounces or inputs a syllable correctly, the system can display text prompts such as "Great job!" or "You're awesome!", or play cheerful sound effects and display cute animated characters. This positive encouragement helps build children's confidence, maintain their enthusiasm for learning, and form a positive learning cycle. Single-key dual-phoneme combinations support visual decomposition display, providing intuitive teaching assistance for single-key dual-phoneme combinations unique to the above methods. For example, when a child learns or inputs a single-key two-phoneme combination, the system can not only display the final syllable, but also break it down into two parts: the medial vowel and the tone, and highlight or animate them separately. It can even play the phonemes of these two parts separately to help the child understand the composition principle and pronunciation logic of the combination, thereby reducing the learning difficulty.

[0279] Through the aforementioned visual teaching functions, this application effectively addresses children's difficulties in mastering pronunciation, lack of interest in learning, and obstacles in understanding complex phoneme combinations in language learning. Animated demonstrations of the mouth shapes for Pinyin pronunciation provide children with intuitive visual references, enabling them to more accurately imitate and correct pronunciation. Three adjustable speech speeds meet the personalized needs of children at different learning stages, ensuring they can learn at their most suitable pace. Interactive syllable-following games transform tedious pronunciation exercises into fun interactive experiences, significantly enhancing children's learning enthusiasm and participation. The encouragement and feedback mechanism after correct input promptly reinforces children's correct learning behaviors and builds learning confidence. Furthermore, the visual breakdown of single-key dual-phoneme combinations greatly simplifies the understanding of complex phoneme structures, allowing children to clearly recognize the components of each phoneme and its pronunciation. The synergistic effect of these functions makes language learning in children's teaching scenarios more vivid, efficient, and effective, significantly improving children's pronunciation accuracy and input efficiency, and cultivating their interest in language learning.

[0280] In multilingual human-computer interaction methods based on number-sequence vowel interfaces, especially when the core support engine is processing user input, users may unintentionally input initial-medial vowel combinations that do not conform to linguistic rules. For example, in Chinese Pinyin input, some initial-vowel combinations do not exist in pronunciation, such as "zh+ü" or "b+iong". If the system simply rejects these invalid inputs without providing clear feedback or guidance, users will find it difficult to understand the reason for the input error, which may lead to repeated attempts and reduced input efficiency, thus affecting the overall user experience.

[0281] To address this, this application further proposes optimizations to the offline encoding verification module of the core support engine. This module can reject invalid combinations such as "zh+ü" and "b+iong". Specifically, the offline encoding verification module of the core support engine is a component of the core support engine, and its main function is to perform real-time verification of user-input initial consonant combinations without relying on a network connection. This module has a built-in comprehensive linguistic rule base, which can identify and reject invalid combinations that do not conform to the pronunciation rules of the target language, such as "zh+ü" or "b+iong" which do not exist in Chinese Pinyin. Through preset rule sets or finite state machines, this module can efficiently determine the validity of input sequences. To ensure the fluency of user input, the response latency of this verification process is strictly controlled within 150ms, which has been verified in a test environment with a room temperature of 25℃ and no external interference. The key to achieving low latency lies in optimizing the verification algorithm, such as using efficient data structures like hash tables or prefix trees for rule matching, and ensuring that the verification module runs with high priority in the system. When validation fails, i.e., an invalid combination is detected, the system does not simply interrupt input. Instead, it prompts the user with a "recommended list of initial consonant combinations" in a fixed order. This recommended list is either predefined or dynamically generated based on the current input context. It contains valid initial consonant combinations that match the already entered parts and is presented in a fixed, user-friendly, and easy-to-understand order (e.g., by frequency of use, alphabetical order, or pronunciation similarity).

[0282] Through the above technical solution, this application effectively solves the confusion and efficiency problems encountered by users when inputting invalid initial-medial consonant combinations. The introduction of the offline encoding verification module ensures that the system can perform real-time and accurate linguistic verification of input even in a network-free environment, thereby preventing invalid combinations from entering the subsequent encoding and candidate processing flow. Rejecting specific invalid combinations such as "zh+ü" and "b+iong" directly improves the accuracy of input. At the same time, controlling the verification response delay to within 150ms ensures that users hardly perceive the verification during the input process, maintaining a smooth interactive experience. More importantly, when input fails, the system no longer simply rejects it, but actively provides a "recommended initial-medial consonant combination list," which provides users with immediate error correction guidance and learning opportunities, significantly reducing the user's trial-and-error costs and frustration, guiding users to quickly input the correct combinations, thereby greatly improving the efficiency of multilingual human-computer interaction and user satisfaction.

[0283] While multilingual human-computer interaction methods based on number-order vowel encoding can achieve multilingual input and prompts in cross-border communication scenarios, users may face language barriers in practical applications, especially when they need to understand and express content in different languages ​​instantly. Traditional input methods typically only provide one-way translation functionality or require switching applications for two-way translation. This not only increases the complexity of operation but also makes it difficult to meet the needs of efficient and seamless cross-language communication, especially when dealing with languages ​​containing tones, where the accuracy and efficiency of input and translation still have room for improvement.

[0284] To address this, this application further proposes supporting bidirectional offline translation between Chinese and English, and languages ​​with tones, in cross-border communication scenarios. This bidirectional offline translation function enables the system to translate between Chinese and English, as well as between Chinese and any language with tones (such as Thai, Vietnamese, etc.), without relying on an external network connection. This is achieved by pre-integrating multilingual bidirectional offline translation functionality and an offline dictionary into the core support engine. The offline translation function ensures that users can still conduct efficient cross-language communication even in environments with poor or no network conditions, improving the system's usability and robustness. To guarantee translation quality, the accuracy rate for commonly used phrases can reach over 92%, and the accuracy rate for specialized terms can reach over 88%. To achieve high accuracy, the system specifically optimizes and expands the offline dictionary for commonly used phrases and specialized terms. This includes, but is not limited to, collecting a large amount of industry corpus, using machine learning models to train and optimize translation algorithms, and regularly updating the dictionary. Through deep learning and matching of specialized terms in specific fields, the system ensures translation accuracy in specific scenarios (such as industrial control, medical, etc.), thereby meeting users' communication needs in different professional backgrounds. Meanwhile, the translation response time has been optimized to within 300ms. To achieve this rapid response, the system optimizes the computational efficiency of the translation algorithm and utilizes the device's local computing resources for offline translation processing. This may involve adopting a lightweight translation model, optimizing data structures and query mechanisms, and accelerating the process at the hardware level. A fast translation response time is crucial for improving user experience, especially during real-time conversations or rapid information retrieval, effectively reducing waiting time and improving communication efficiency.

[0285] Furthermore, the single-key dual-phoneme combination rule is reused for tonal languages. A single-key dual-phoneme combination refers to an input method that triggers two phonemes with a single keystroke, such as the combination of a medial vowel and a tone. Reusing this for tonal languages ​​means that when inputting these languages, users can efficiently input syllables with tones using a single-key dual-phoneme input logic similar to Pinyin. This requires the system to reserve encoding positions for tonal languages ​​in the numerical vowel encoding system and associate the phonemes of tonal languages ​​with numerical vowels and the single-key dual-phoneme combination rule through speech feature mapping, thereby achieving unified and efficient input. In bilingual mixed-tone input, the conflict rate is controlled below 3%. Bilingual mixed-tone input means that in the same input session, users can simultaneously input two or more tonal languages ​​without frequently switching input modes. To reduce the conflict rate, the system utilizes its multilingual unified encoding logic, based on the numerical vowel core encoding rule, to achieve unified encoding of different languages ​​through "vowel reserved encoding positions + speech feature mapping". In addition, the encoding module's built-in encoding conflict detection unit monitors and resolves potential encoding conflicts in real time, ensuring that the system can accurately identify and generate the correct toned syllables or characters when mixed input is used, thereby providing a smooth and error-free input experience.

[0286] Through the aforementioned technical solutions, this application significantly improves the efficiency and accuracy of multilingual human-computer interaction in cross-border communication scenarios. Users can achieve real-time bidirectional translation between Chinese and English and toned languages ​​without relying on the network, effectively solving communication barriers in environments with limited network coverage. High-accuracy translation of common phrases and professional terms, combined with rapid response time, greatly optimizes the fluency and professionalism of cross-language communication. Furthermore, the single-key dual-phoneme combination rule is reused for toned languages, enabling users to input complex syllables with tones in a more intuitive and efficient way, reducing input steps and cognitive burden. Simultaneously, through optimized multilingual unified encoding logic and conflict detection mechanisms, an extremely low conflict rate is ensured when inputting mixed bilingual tones, thus providing users with a seamless, efficient, and accurate cross-language human-computer interaction experience, especially when handling complex mixed input of toned languages, where its advantages are even more pronounced.

[0287] In multilingual human-computer interaction methods based on number-sequence vowel interfaces, this application proposes various scenario-customized adaptation schemes to achieve cross-device adaptation, multi-directional layout adaptation, and core engine support. However, in industrial control scenarios, higher demands are placed on the input efficiency of core commands, offline availability, and the rapid and accurate input of complex commands with tuning. Traditional input methods may suffer from low input efficiency and error-proneness due to unstable network environments or command complexity, thereby affecting the continuity and safety of industrial production.

[0288] In response, this application further proposes an optimization scheme for industrial control scenarios, which includes: supporting offline caching of core instructions for industrial control scenarios, with a cache quantity of no less than 1,000 instructions and a retention period of no less than 7 days, ensuring that the input time of core instructions does not exceed 1 second; at the same time, after integrating single-key dual-phoneme combinations, the input time of instructions with tone does not exceed 180ms / instruction, the offline cache combination instructions are no less than 800 sets, and the availability after network outage is no less than 99.8%.

[0289] This solution aims to ensure fast and reliable input of critical commands in industrial control scenarios. Offline caching refers to pre-storing frequently used or operationally critical commands on local devices, making them accessible and usable even in environments without network connectivity or with unstable networks. Specifically, the system can allocate a dedicated area in the device's non-volatile memory to store these core commands. The cache management mechanism is responsible for command synchronization, updating, deletion, and expiration handling, ensuring that the number of cached commands is no less than 1000 and their retention period is at least 7 days to meet the needs of long-term offline operations. The input time for core commands should not exceed 1 second, meaning that the entire process from user-triggered input to system recognition and preparation for command execution must be highly efficient. This can be achieved by optimizing command retrieval algorithms (e.g., using hash tables or B-tree index structures), reducing intermediate processing steps, and preloading frequently used commands into memory.

[0290] Furthermore, this solution further improves the input efficiency and reliability of complex instructions in industrial control scenarios, especially for instructions requiring tone differentiation. Single-key dual-phoneme combinations allow users to input complex phonemes containing both vowels and tones with a single key press, significantly reducing input steps. Integrating this function into industrial control scenarios means the system can recognize and process these encoded sequences generated by single-key dual-phoneme combinations and map them to corresponding tone-based instructions. To ensure that tone-based instruction input takes no more than 180ms per instruction, the system needs to optimize the parsing speed of single-key dual-phoneme encoding and the matching efficiency with offline cached instructions. The offline cached instruction set should contain at least 800 sets, requiring the system to store a large number of such complex instructions and their corresponding single-key dual-phoneme encodings to meet the needs of different industrial processes. A minimum offline availability of 99.8% emphasizes that even in a completely offline state, the input and recognition functions of these tone-based instructions must maintain extremely high reliability. This is typically achieved by completely localizing the single-key dual-phoneme encoding parsing logic, instruction matching algorithm, and cached data, and by conducting rigorous offline testing and redundant design.

[0291] Through the above technical solutions, this application effectively solves the problems of core command input efficiency and offline availability in industrial control scenarios. Firstly, by supporting offline caching of core commands, operators can quickly and reliably input critical commands even in industrial environments with unstable or no network connection, ensuring the continuity and safety of the production process. A large number of cached commands and a long retention period ensure that frequently used commands are always readily available, significantly reducing input obstacles caused by network latency or interruptions. Secondly, by integrating single-key dual-phoneme combination functionality and combining it with offline caching, the input efficiency and accuracy of commands with tones are greatly improved. This allows operators to quickly input complex commands with specific semantic distinctions with fewer keystrokes, especially suitable for command scenarios requiring precise pronunciation or semantic differentiation. Simultaneously, offline caching of combined commands and extremely high offline availability further guarantee the efficiency and reliability of complex command input in harsh industrial environments, thereby comprehensively improving the human-machine interaction efficiency and system robustness in industrial control scenarios.

[0292] In some embodiments described above, this application proposes a multilingual human-computer interaction method based on a number-order vowel interface. This method achieves unified encoding logic for multiple languages ​​by defining a unique set of number-order vowels and mapping rules, configuring cross-device compatible numeric keys, constructing an irreplaceable number-order vowel interface, and establishing a collaborative parent-child interface system. However, in practical applications, for languages ​​with tones such as Chinese, inputting a combination of medial vowel and tone, or a combination of tone and medial vowel, may require multiple key presses. This affects input efficiency and fluency to some extent, especially when rapid input of specific syllable combinations is required, where the user experience may be less than ideal.

[0293] To address this, this application further proposes a single-key dual-phoneme enhanced human-computer interaction method, aiming to optimize the input efficiency of medial vowel and tone combinations. This method first defines single-key dual-phoneme combination types, mainly including combinations of medial vowels with number-order single tones or character tones (e.g., the medial vowel i with 1-4 tones / neutral tone, or the medial vowels u / ü with 1-4 tones / neutral tone), and combinations of tones with medial vowels (e.g., the 1-4 tones / neutral tone with the medial vowel i, or the 1-4 tones / neutral tone with the medial vowel u / ü), where the order of tones can be adjusted according to actual needs. These combinations represent common phoneme structures in Chinese that require efficient input.

[0294] To achieve single-key triggering of these dual-phoneme combinations, this method provides two irreplaceable implementation paths. The first is long-press numeric key triggering, where a long press of a specific numeric key (e.g., long press of the 0 key is associated with the vowel 'i', long press of the "-" key is associated with the vowels 'u' and 'ü') automatically identifies and outputs the corresponding dual-phoneme combination when the press duration reaches or exceeds a preset threshold (e.g., ≥300ms, user-defined threshold range 100-500ms). The second is shortcut key combination triggering, implemented through preset key combinations. These key combinations are locked via system permission requests to prevent third-party software from using them, ensuring that triggering is directly responded to through the system's underlying interface without being intercepted by third-party software. In this method, dual-key input is triggered by short presses (e.g., short press duration ≤200ms). The system distinguishes different triggering methods through a dual determination mechanism of press duration and signal characteristics, thus avoiding conflicts and ensuring a trigger response delay ≤50ms. Furthermore, this method emphasizes that the encoding sequence of single-key diphone combinations has a unique correspondence with the core encoding system of number-order vowels and cannot be equivalently replaced. The aforementioned "single-key triggering-encoding mapping-cross-mode compatibility" constitutes an indivisible functional whole, and it is prohibited to circumvent this by splitting "triggering operation and encoding generation" into independent steps. That is, any scheme that first triggers the key and then generates an equivalent diphone encoding through an additional software module falls within the protection scope of this method.

[0295] Furthermore, this method ensures cross-mode compatibility of the single-key dual-phoneme combination rules, making it applicable to the tonic set, consonant set, and all Pinyin input modes, and enabling it to work in conjunction with the dual-mode tone function. This means that regardless of the user's input mode, this efficient single-key dual-phoneme input mechanism can be used, greatly improving the flexibility and convenience of input.

[0296] Through the above technical solution, this application effectively solves the problem of low input efficiency for specific phoneme combinations such as medial vowels and tones in multilingual human-computer interaction methods based on number-order vowel interfaces. Users can quickly and accurately input syllables containing both medial vowels and tones (i.e., tone-marked medial vowels) by simply pressing a single key or using shortcut key combinations, without needing to perform multiple key presses or complex combination inputs. This significantly improves input efficiency and fluency. Especially for languages ​​with tones such as Chinese, this single-key dual-phoneme input method allows complex syllable structures to be input more intuitively and quickly, reducing the user's cognitive burden and operational difficulty. Simultaneously, the direct response and dual-judgment mechanism of the system's underlying interface ensures accurate triggering and low latency, providing users with a more stable and reliable input experience. Furthermore, the indivisible design of this solution effectively guarantees the integrity and uniqueness of the core technical solution, preventing circumvention through partial implementation, thus ensuring that the technical advantages brought by this method can be fully utilized. Its cross-mode compatibility also allows this efficient input mechanism to be widely applied in various input scenarios, further enhancing the practicality and user-friendliness of the entire interactive system.

[0297] While single-key dual-phoneme combination input based on the numerical vowel core encoding system can effectively improve input efficiency, a single or fixed tone triggering method may not fully adapt to different user habits and diverse input scenarios. For example, users accustomed to number input may prefer to trigger tones directly via number keys; while users accustomed to traditional keyboard layouts or needing to quickly input punctuation may prefer tone triggering combined with punctuation input to reduce key switching. This single tone triggering mechanism may lead to a poor user experience and reduced input flexibility.

[0298] In this regard, this application further proposes optional tone triggering methods for single-key dual-phoneme combinations, specifically including: single-tone numbers triggered by numbers 1-4 and 5, dual-tone numbers triggered by numbers 6 and 7-0, and character tones triggered by independent keys or "punctuation-tone" shared keys.

[0299] The feature "selectable tone triggering method for single-key dual-phoneme combinations" means that when performing single-key dual-phoneme input, users can choose the most convenient tone input method based on their personal preferences, device type, or current input scenario, rather than being limited to a single fixed method. This selectivity aims to improve the system's flexibility and user-friendliness.

[0300] "Single tone triggered by numbers 1-4 and 5" is a tone triggering method where the four tones of Mandarin Chinese (first, second, third, and fourth tone) as well as the neutral tone or no tone are mapped to the number keys 1 through 5. Specifically, the number 1 triggers the first tone, number 2 triggers the second tone, number 3 triggers the third tone, number 4 triggers the fourth tone, and number 5 is used to trigger the neutral tone or no tone. This method utilizes the intuitiveness and universality of number keys, allowing users to quickly add tones directly using the number keys after inputting the initial-medial-vowel combination. It is especially suitable for users accustomed to number input or operating on a numeric keypad layout.

[0301] "Numerical double tone triggering via numbers 6, 7-0" is another tone triggering method that extends the functionality of numerical single tone by providing a second set of tone configurations. In this configuration, the number 6 can correspond to a neutral tone or no tone, while the numbers 7 through 0 correspond sequentially to the four tones of Mandarin Chinese (e.g., 7 corresponds to the fourth tone, 8 to the first tone, 9 to the second tone, and 0 to the third tone), and this order can be customized. By providing two sets of numerical tone configurations, the system can simultaneously prompt two sets of candidate syllables with tones, thus offering richer candidate choices in certain complex contexts and allowing users to personalize settings based on high-frequency tones or personal habits.

[0302] "Character tone triggering via dedicated keys or 'punctuation-tone' shared keys" is a method that combines tone triggering with character keys (especially punctuation keys). "Dedicated key triggering" refers to setting up a dedicated, independent key in the keyboard layout to trigger the tone. These dedicated keys are usually the same size as the numeric keys and can adapt to different layouts, such as a 4x4 or circular layout, providing users with a clear and unambiguous tone input entry point. "'Punctuation-tone' shared key triggering" refers to multiplexing the tone function with commonly used punctuation keys (such as commas, periods, question marks, exclamation marks, etc.). For example, a short press outputs a punctuation mark, while a long press or key combination triggers the tone. This method effectively saves keyboard space, especially suitable for small-screen devices or scenarios requiring a compact layout, while maintaining input efficiency.

[0303] By providing multiple tone triggering methods with single-key dual-phoneme combinations, this application effectively solves the limitations of a single tone triggering mechanism in adapting to user habits and diverse input scenarios. Specifically, users can flexibly choose to trigger a single tone using numbers 1-5, or a dual tone using numbers 6, 7-0, or a character tone using a separate key or a shared "punctuation-tone" key, based on their preferences and device characteristics. This selectivity greatly enhances the flexibility of human-computer interaction and user experience. For example, users accustomed to number input can quickly add tones using number keys, reducing learning costs and operation steps; while users requiring a compact layout or accustomed to traditional keyboards can achieve tone input without adding extra keys using the shared "punctuation-tone" key, effectively saving screen space. Furthermore, the introduction of dual-tone numbers, by providing a second set of tone configurations and custom sorting functions, enables the system to simultaneously prompt richer candidate syllables with tones, further improving the accuracy of candidate words and input efficiency. This diverse tone triggering mechanism, combined with the aforementioned single-key dual-phoneme enhanced human-computer interaction method, enables users to input tone-infused syllables in a more intuitive and efficient manner when inputting combinations of medials and tones, thereby significantly improving overall input efficiency and user satisfaction.

[0304] While the single-key dual-phoneme enhanced human-computer interaction method described above significantly improves Pinyin input efficiency, especially when handling medial and tone combinations, through its unique single-key triggering logic and encoding mapping, its application in professional fields such as medicine, finance, and industry faces challenges. These fields contain numerous unique, high-frequency technical terms, often including specific tone combinations. A generalized dual-phoneme combination rule base may not efficiently and accurately identify and prompt these technical terms, requiring professional users to spend additional time selecting or correcting them during input, thus impacting input efficiency and accuracy in professional scenarios.

[0305] In response, this application further proposes support for industry-specific extensions. Specifically, the medical, financial, and industrial sectors can import their own dual-phoneme combination rule bases, and the system will automatically associate industry-specific tone combinations; the input response latency for professional terms with tone is ≤70ms, and the annotation accuracy is ≥99%.

[0306] The ability to support industry-specific extensions means that this method can be customized to meet the needs of different industries, adapting to the input habits and vocabulary characteristics of specific professional fields. Specifically, the system can be designed with a modular architecture, allowing the loading or activation of industry-specific input rules, lexicons, and recognition models. For example, in the medical field, a dedicated rule base containing a large number of medical terms, drug names, and disease diagnoses can be loaded; in the financial field, a rule base containing financial products, market terminology, and regulatory provisions can be loaded. This scalability ensures that the input method, while maintaining general applicability, can also accommodate the specificities of professional fields.

[0307] In the medical, financial, and industrial sectors, proprietary dual-phoneme combination rule bases can be imported, providing a concrete way to achieve industry-specific extensions. A proprietary dual-phoneme combination rule base refers to a database that is pre-collected, organized, and optimized for a specific industry, containing high-frequency professional terms for that industry and their corresponding tone-encoded dual-phoneme sequences. The import process can be configured through the system management interface or integrated with industry application systems via API interfaces. For example, in the medical field, tone-encoded dual-phoneme sequences for professional terms such as "myocardial infarction" and "hypertension" can be imported; in the financial field, sequences for terms such as "equity pledge" and "quantitative trading" can be imported. These rule bases coexist with general terminologies and are prioritized for use under specific industry models.

[0308] The system automatically associates industry-specific tone combinations. After importing a dedicated rule base, the system can intelligently identify and associate user-inputted diphone sequences with industry-specific tone combinations. When a user inputs in a specific industry mode, the system prioritizes matching professional terms from the dedicated rule base. For example, when a user inputs a combination of a medial vowel and a tone, the system will automatically associate this combination with preset professional terms in the currently active industry rule base and suggest them as high-priority candidates. This automatic association mechanism reduces the number of manual selection steps for users and improves the efficiency of professional terminology input.

[0309] The system requires a response latency of ≤70ms for inputting technical terms with tones. Response latency refers to the time interval between the user completing the input of a technical term with tones and the system displaying the corresponding candidate word. To achieve a low latency of ≤70ms, the system needs to employ efficient data retrieval algorithms and optimized memory management mechanisms to ensure rapid feedback even when querying a large, proprietary rule base. For example, data structures such as hash tables or B-trees can be used to store the rule base, and multi-threading technology can be used to execute query operations in parallel, thus ensuring a smooth experience for professional users even during high-speed input.

[0310] A labeling accuracy rate of ≥99% is a quantitative requirement for the system's accuracy in recognizing technical terms. Labeling accuracy refers to the proportion of times the system can correctly match and indicate the target technical term when recognizing the diphone combination with tone in technical terms. To achieve a high accuracy rate of ≥99%, the construction of the dedicated rule base requires rigorous corpus training and manual proofreading to ensure that the diphone encoding with tone for each technical term is completely consistent with the actual pronunciation and writing. Furthermore, the system can integrate context analysis and semantic understanding modules to further improve the accuracy of technical term recognition in complex contexts and reduce misjudgments and ambiguities.

[0311] Through the above technical solution, this method effectively addresses the problem of insufficient recognition and input efficiency of high-frequency professional terms in general input methods within specialized fields. By supporting industry-specific extensions and allowing the import of dedicated dual-phoneme combination rule bases for fields such as medicine, finance, and industry, the system can specifically optimize the recognition and prompting of professional terms. When users input in specific professional scenarios, the system can automatically associate and prioritize industry-specific tone combinations, greatly reducing the time professional users spend manually searching and correcting. Simultaneously, the response latency for inputting professional terms with tone is controlled within 70ms, ensuring a smooth experience for professional users during high-speed input and avoiding efficiency drops due to system lag. Furthermore, the high annotation accuracy rate of up to 99% significantly reduces the input error rate of professional terms, reducing the burden of proofreading and modification for users, thereby comprehensively improving the efficiency and accuracy of human-computer interaction in professional fields. This makes the single-key dual-phoneme enhanced human-computer interaction method more practical and competitive in professional applications.

[0312] In some of the embodiments described above in this application, a multilingual human-computer interaction method based on the number sequence vowel interface is proposed, and further functions such as number sequence tone expansion and single-key dual-phoneme enhancement are provided. However, in practical applications, users often have diverse input habits and scenario requirements. A single or limited input mode may not be able to fully meet the personalized needs of all users, thereby affecting input efficiency and user experience.

[0313] To address this, this application further proposes a multi-input mode human-computer interaction method, aiming to improve overall input efficiency and user satisfaction by providing multiple input modes to adapt to different users' input habits and scenario needs. The method includes:

[0314] First, it achieves full coverage of input modes, including six modes: abbreviated pinyin, abbreviated pinyin extension, overlapping pinyin, moving double pinyin, single-key full pinyin, and double-key full pinyin. All these modes are built upon the aforementioned numerical vowels, ensuring the uniformity and compatibility of the underlying encoding logic.

[0315] Secondly, core functionalities are provided for each mode.

[0316] The simplified Pinyin mode supports encoding of Chinese and English initials / core syllables as well as bilingual mixed encoding, enabling fast input with an index response latency of less than or equal to 200ms. The simplified Pinyin mode can reserve encoding space for up to 6 columns × 11 = 66 multilingual extended characters, allowing input of Chinese, Chinese Pinyin and words / phrases or both, Chinese-English bilingual, or Chinese-English and multilingual.

[0317] The simplified pinyin extension mode adds the logic of "medial vowel + single-key number-order vowel" to the simplified pinyin logic. The triggering method of medial vowel is flexible and diverse. Users can choose to long press the soft keyboard to pop up the window, slide the direction, or long press the hard keyboard to activate and then short press to output, in order to adapt to different devices and operating habits.

[0318] It pioneered a single-key dual-phoneme combination mode for overlapping and tone marking of medial vowels. The first and second images of each pair of images have two different input modes for initial and medial vowels, including a second medial vowel input method, which is completely different from the conventional medial vowel input method. After mastering it, you can type blindly: Images 6-7, 9-10, 12-13, 15-16, 18-19, 20-21, 22-23, 25-26, 28-29, 31-32, and 33=34, which respectively use the dual-key triggering method for initial and medial vowels. It is compatible with abbreviated spelling, abbreviated spelling extension, and full spelling modes, supports mutual conversion between modes, and has an accuracy rate of greater than or equal to 99%. After the initial consonant is entered, up to nine initial consonants and English characters containing the initial consonant are displayed. Only the 0 and 11 keys are available. The 0 key (bound to the medial i) and the "-" key (bound to the medial u / ü) can be used for single-key double-phoneme combinations, i.e., tone marking medial consonants. As a single-key double-phoneme combination, it can be entered with a single keystroke before or after the input of the numerical vowel. This is the first of its kind in this application, while the existing technology requires two keystrokes. This improves the input efficiency of tone and medial consonant by 50%.

[0319] Figures 49-52 illustrate examples of mobile double-pinyin input modes with four different numerical vowel layouts and the same tone-marking medial layout. These modes are optimized for mobile devices and are compatible with Chinese-English bilingual or multilingual input modes. Lowercase / uppercase initials and English characters, along with 3 columns × 11 = 33 multilingual extended characters (exclusive to the overlapping pinyin mode), are all triggered by a double key. Selecting uppercase initials and English characters enters the English input mode, while selecting lowercase initials enters the Chinese input mode. For medial vowels with zero medial vowel, the numerical vowel layout is used directly. This mode reuses the numerical dual-tone system from the aforementioned numerical tone extension method and the aforementioned single-key dual-phoneme combination rules. After inputting the initial consonant and the numerical vowel, the system automatically triggers a single-key dual-phoneme combination via the 0 key (bound to the medial vowel i) or the "-" key (bound to the medial vowels u / ü), directly prompting the simplified code or syllable symbol of the initial-medial-vowel-tone syllable containing the medial vowel. The combination response latency is less than or equal to 50ms (test environment: room temperature 25℃±5℃, device memory greater than or equal to 2GB). For simplified code words or syllable symbols of the initial-medial-vowel-tone syllable without the medial vowel, it supports input via the above-mentioned consonant set character tone (including independent tone keys or "punctuation-tone" shared key keys), or reuse of the above-mentioned numerical tone system input. Users can choose freely, and both methods share the tone-matched vocabulary mapping library. Specifically, the 0 key corresponds to i + numerical vowel + tone, while the "-" key corresponds to u / ü + numerical vowel + tone. Furthermore, the mobile double-pinyin mode incorporates a splitting avoidance identification logic. For splitting behaviors such as "splitting the initial consonant input and the medial consonant trigger as two independent software calls" or "splitting the encoding generation and candidate prompts as different modules," it automatically identifies and rejects equivalent inputs by detecting the temporal correlation of the input signals (the interval between consecutive operations within the same input session is less than or equal to 100ms) and the module call chain. Simultaneously, a splitting avoidance verification standard is clearly defined: if the input steps, encoding generation logic, and candidate output results of a certain scheme are more than or equal to 90% equivalent to the core features of this mode, and are achieved through splitting technical features, then it is determined to be an infringement.

[0320] The single-key full-pinyin mode is mainly used to adapt to existing technology comparison scenarios, allowing users to input pinyin characters one by one.

[0321] The two-key full-spelling mode corresponds to the single-key full-spelling logic of the letter keyboard, and is especially suitable for children's basic language teaching scenarios.

[0322] Finally, the method also supports mode cooperative switching, that is, it supports automatic switching based on the input sequence or manual switching by the user, with a switching delay of less than or equal to 50ms.

[0323] Through the above technical solutions, this application provides a comprehensive and flexible multi-input mode system, effectively resolving the contradiction between diverse user input needs and a single input mode. The introduction of various modes such as simplified pinyin, simplified pinyin extension, overlapping pinyin, mobile double pinyin, single-key full pinyin, and double-key full pinyin allows users to freely choose the most suitable input method according to their personal habits, device type, and input scenario, greatly improving input efficiency and user experience. In particular, the mobile double pinyin mode, by reusing number sequence tone extension and single-key dual-phoneme combination rules, achieves efficient single-key input of medials and tones, significantly reducing the number of keystrokes and accelerating input speed. Its built-in splitting avoidance recognition logic further ensures the integrity and security of the core encoding system, preventing malicious splitting and avoidance behaviors. The collaborative switching function between modes, whether automatic or manual, ensures the smoothness and continuity of the input process, avoiding the sense of interruption caused by mode switching. Overall, this method, while maintaining the uniformity of the underlying number sequence vowel encoding, provides rich upper-level input interaction options, making human-computer interaction more intelligent, efficient, and personalized.

[0324] In some of the embodiments described above in this application, the simplified pinyin extension mode supports triggering the medial vowel by superimposing the logic of the medial vowel and the single-key numerical vowel. However, in practical applications, if there is a lack of refined definition and real-time feedback mechanism for these triggering methods, users may face problems of operational uncertainty and low efficiency when inputting the medial vowel, especially in scenarios where it is necessary to input syllables with tones quickly and accurately. This uncertainty will affect the overall input fluency.

[0325] To address this, this application further proposes a triggering rule for the soft keyboard in the simplified pinyin extension mode. Specifically, when a user operates on the 0 key or the "-" key on the soft keyboard, the system detects the press duration. If the press duration reaches or exceeds a preset threshold (e.g., 300 milliseconds for regular users and 500 milliseconds for children's mode users), a large pop-up window containing candidate pins will appear on the screen. This long-press triggering mechanism based on a precise duration threshold effectively avoids accidental touches, while the large pop-up window provides clear visual feedback, facilitating user selection among multiple candidate pins. This significantly improves the accuracy and convenience of operation, especially on devices with smaller screen sizes. Alternatively, users can trigger pin input by directional sliding on the pin keys. For example, starting from the 0 key or the "-" key, sliding in a specific direction (such as up, down, left, or right) will allow the system to identify and input the corresponding pin or pop up relevant candidates based on the sliding direction. This gesture operation provides a more flexible and faster input method, especially suitable for users accustomed to swipe input, reducing the frequency of fingers leaving the keyboard area and improving input consistency. Regardless of whether a long press pop-up or directional swipe is used, the system's recognition accuracy for user actions must reach over 96%. This means the system can accurately determine the user's intention, effectively distinguishing between different operations such as long press, short press, and swipe, thereby ensuring the accuracy of medial input and reducing input efficiency and user frustration caused by recognition errors. When a user successfully inputs a medial using any of the above triggering rules, the system will immediately and synchronously display the associated "medial + tone" single-key dual-phoneme combination on the interface. For example, if the medial "i" is triggered, it may display tone-included syllable combinations such as "i1 / ī", "i2 / í", "i3 / ǐ", "i4 / ì", and "i5 / i". This instant and intuitive feedback mechanism not only confirms the user's input but also pre-displays possible tone-included syllable options, greatly simplifying the subsequent tone input process and improving the efficiency and accuracy of tone-included syllable input.

[0326] Through the above technical solution, this application provides clear and efficient triggering rules for soft keyboard input of medial vowels in the simplified pinyin extended mode. By precisely controlling the long press duration or using directional swiping, users can significantly improve the accuracy of medial vowel triggering, avoiding accidental touches and repeated inputs caused by ambiguous operations. In particular, the design of a large candidate pop-up window after a long press provides clearer visual selection on small-screen devices, reducing the difficulty of selection. At the same time, a recognition accuracy rate of up to 96% ensures accurate capture of the user's operational intent. More importantly, the simultaneous display of the "medial vowel + tone" single-key dual-phoneme combination greatly simplifies the input process of syllables with tones. Users do not need to trigger the tone key again to directly select the desired syllable with tone, thereby significantly improving the input efficiency and user experience of syllables with tones in the simplified pinyin extended mode, making the input of complex syllables smoother and more intuitive.

[0327] In some of the above implementations, a multilingual human-computer interaction method based on a numerical vowel interface is proposed. Its mobile double-pinyin mode achieves efficient pinyin input by reusing the numerical tone system and single-key dual-phoneme combination rules. However, for skilled users accustomed to touch typing on a physical keyboard, the default key combination order may not fully match their personalized input habits and muscle memory, thus affecting input efficiency and accuracy. If it cannot be adjusted according to user preferences, it may limit the application potential and input experience of this mode among professional users.

[0328] In response, this application further proposes rules for blind typing on a mobile double-keyboard hard keyboard, which allows users to adjust the combination order; skilled users can achieve a blind typing accuracy of ≥95%, and it supports binding single-key dual-phoneme combination shortcut keys with a response latency of ≤15ms.

[0329] Specifically, supporting user-adjusted key combination order means the system provides a mechanism that allows users to customize the key combination order of the physical keyboard in mobile double-pinyin mode according to their personal habits or preferences. For example, users can drag or remap specific initial consonants, finals, or medials with tones through the system settings interface to optimize their finger path and keying habits when touch typing. This adjustment can include optimizing the key sequence of specific syllable combinations or mapping frequently used combinations to more convenient positions, thereby improving input efficiency and comfort. A touch typing accuracy rate of ≥95% for skilled users means that after user-customized adjustments, skilled users who have used this input method for a long time can achieve an accuracy rate of 95% or higher when typing without looking at the keyboard. Achieving this high accuracy rate relies not only on the user's mastery of the customized key combination order but also on the system's precise recognition and error-tolerant processing of subtle operations such as key timing and pressure, ensuring accurate recognition of the user's intent even during rapid touch typing. Support for binding single-key dual-phoneme combination shortcuts means that this function allows users to bind single-key dual-phoneme combinations, such as "i + 1st tone" or "u + 2nd tone," to specific shortcut keys or key combinations on the physical keyboard. Users can select one or more physical keys through the system's shortcut key management interface and configure them to trigger a specific single-key dual-phoneme combination. For example, a less frequently used function key can be set as a shortcut key for "i + 1st tone." When the user presses this key, the system directly outputs the corresponding toned syllable without requiring multiple key presses. A response latency of ≤15ms means that the time interval between the user pressing a key on the physical keyboard and the system recognizing and processing the input signal, and preparing to output or display the candidate result, does not exceed 15 milliseconds. Such low response latency is crucial for touch typing users, ensuring smooth input and immediate feedback, avoiding input interruptions or misjudgments caused by system lag, thereby greatly improving the user's input experience and efficiency.

[0330] By introducing personalized configuration of touch typing rules for mobile dual-pinyin hard keyboards, this application effectively solves the problem of insufficient adaptability of skilled users to the default combination order when inputting on a hard keyboard. Users can adjust the combination order according to their own habits and bind frequently used single-key dual-phoneme combination shortcut keys. This directly optimizes the finger path, reduces unnecessary key operations, and thus significantly improves the smoothness and efficiency of touch typing. At the same time, the system's ultra-low response latency to input operations ensures that users can get instant feedback when touch typing at high speed, avoiding input interruptions or errors caused by latency, and greatly improving the touch typing accuracy of skilled users. This personalized and highly responsive hard keyboard input experience makes the mobile dual-pinyin mode more practical and competitive in professional and high-efficiency input scenarios, giving full play to the advantages of the number sequence vowel encoding system.

[0331] This application proposes a multilingual human-computer interaction method based on a number-sequential vowel interface. This method aims to achieve efficient and unified cross-language input through a unique encoding system and interface design. However, when applying this innovative input method to diverse devices and complex scenarios, ensuring the immutability of core encoding rules, the security of the input process, seamless collaboration between different modules, and effective prevention of potential circumvention behaviors are key challenges to achieving its stable and reliable operation.

[0332] In this regard, this application further proposes a multilingual human-computer interaction system, which includes an input module, an encoding module, a candidate processing module, a display module, a control module, and a hardware encryption unit.

[0333] The input module, serving as the first line of defense in human-computer interaction, is responsible for capturing user actions on different physical or virtual interfaces and converting the user's intent into signals recognizable by the system. Specifically, this module receives keystrokes from either a soft or hard keyboard. These actions include not only key identifiers but also physical characteristics such as press duration and pressure, used to distinguish between short presses, long presses, or to filter out accidental touches. Simultaneously, it can capture the user's voice through a microphone and perform preliminary acoustic feature extraction, such as voiceprint recognition and syllable feature analysis. Furthermore, this module can recognize the user's gesture trajectories and patterns on touchscreens, in the air, or on specific sensors, such as combinations of swipes, pinches, and clicks. Finally, the input module transforms these raw inputs into standardized input signals with feature identifiers for subsequent processing by later modules.

[0334] The encoding module is the core logical unit of the system, responsible for converting user input received by the input module into internal encoding. This ensures that the input signal can be accurately and uniquely mapped to a predefined encoding sequence and prevents invalid combinations. This module pre-stores unmodifiable number sequence vowel mapping rules and parent-child interface configurations. These rules and configurations are fundamental to the system's operation; they are embedded within the module and protected by a hardware chip-level encryption unit to prevent any unauthorized modification. These rules correspond to the methods described in any one of claims 1-26 above, such as the three-in-one mapping table defined in claim 1 and the parent-child interface system configuration in claim 1. Input signals with feature identifiers from the input module are parsed by the encoding module and converted into corresponding initial-medial-final codes or multilingual character codes according to the preset mapping rules. This module incorporates a built-in encoding conflict detection unit. During the encoding conversion process, it verifies the validity of generated encoding combinations in real time. For example, for Chinese Pinyin, it rejects combinations such as "zh+ü" and "b+iong" that are linguistically invalid, ensuring the accuracy and validity of the encoding. The response delay for rejecting invalid encoding combinations is ≤150ms (test environment: room temperature 25℃±5℃, humidity 40%-60%, no external electromagnetic interference). The encoding module is tightly integrated with the hardware chip-level encryption unit, ensuring that the core mapping rules and configurations cannot be modified by external software or unauthorized users at both the physical and logical levels. This protects the core patented technology of the input method from being circumvented or counterfeited.

[0335] The candidate processing module is responsible for generating a list of selectable characters or words based on the encoding output by the encoding module. This aims to provide intelligent and efficient candidate suggestions to help users quickly complete input. The module calls an encrypted offline dictionary based on the encoding. This dictionary contains a large number of words, phrases, and characters, and is associated with ordinal vowel encodings. Based on the input encoding sequence, it retrieves all matching characters or words from the offline dictionary and generates a candidate list. To improve user selection efficiency, the candidate list supports dual sorting by input frequency and semantic relevance, placing the most likely words to be selected first. The update latency of the candidate list is ≤30ms, ensuring that users can quickly see the candidate list after input, providing a smooth input experience.

[0336] The display module is responsible for presenting various interfaces and information of the system to the user, providing clear, intuitive, and adaptive visual feedback. This module can display the vowel interface, candidate list, and extended interfaces. Depending on the current input mode and device type, it displays the number vowel keyboard interface, candidate word list, and other extended function interfaces. To optimize user experience, interface rendering follows the principle of "horizontal alignment of the trigger area and candidate area." The candidate list is typically displayed near the area where the user triggers the input operation and remains horizontally aligned, reducing the distance the user's eyes need to move and improving input efficiency. The alignment error is ≤1 pixel. Considering that this method needs to adapt to various devices, from smartwatches to industrial control panels, the display module supports multi-device screen size adaptation (adaptation range: 1.5 inches - 21 inches). It can automatically adjust the interface layout, key size, and font size according to the device's screen size, ensuring a good visual effect and operating experience on screens of different sizes.

[0337] The control module is the central nervous system of the system, responsible for coordinating the work of various modules, ensuring the coordinated operation of all parts of the system, realizing complex functions, and maintaining the integrity and security of the system. This module adopts a multi-threaded collaborative architecture to handle concurrent user input, background encoding, dictionary lookup, and interface rendering tasks, enabling parallel execution of each task and improving system response speed and overall performance. It controls the synchronous operation of the aforementioned modules, coordinating the data and control flows between the input module, encoding module, candidate processing module, and display module, ensuring they work collaboratively according to predetermined logical order and time requirements. This module implements input mode switching, multimodal interaction, and cross-device adaptation. Based on user operation or system status, it can smoothly switch between different input modes, manage the integration of multiple interaction methods such as voice and gestures, and make corresponding adaptation adjustments according to device type, with a switching latency of ≤50ms. Crucially, the control module has a built-in anti-circumvention monitoring unit, a key security feature designed to protect the core technology of this method from malicious disassembly or circumvention. This unit identifies and blocks circumvention behaviors using the following rules: First, it monitors the module call chain, checking the call relationships between modules within the system (such as encoding conversion, conflict detection, and candidate generation). If a scheme is found to split these closely related functions into independent modules and lacks collaborative verification, it is identified as circumvention behavior and blocked. Second, it detects the temporal continuity of input signals. Core operations should normally be continuous and closely related. If a scheme is found to simulate core functions through discontinuous operations, it will be blocked. Third, it verifies the integrity of core technical features, checking whether the system has completely called all necessary modules. If it detects that only some modules are called (such as only the input and display modules, skipping the encryption verification of the encoding module) to achieve equivalent input, it directly rejects and records the circumvention behavior. Finally, for small-screen devices such as smartwatches, it blocks circumvention behaviors such as "using only '11' or '12' extended keys but deviating from the core mapping table" or "modifying the key-vowel correspondence of small-screen devices."

[0338] The hardware encryption unit provides physical-level security, ensuring the confidentiality and integrity of core data and rules, and preventing unauthorized access, tampering, or reverse engineering of core technical assets. Integrated with the encoding module, this unit provides hardware-level encrypted storage for core mapping rules and offline dictionaries. These core assets are stored within the hardware encryption unit and protected by robust encryption algorithms, ensuring that even if the device is physically accessed, this data cannot be easily read or modified. The unit only allows reading through authorized interfaces and prohibits modification. Only authorized system components can access this encrypted data through specific secure interfaces, and access permissions are strictly limited to read-only, fundamentally eliminating the possibility of tampering. The encryption algorithm adopts both the Chinese national standard SM4 and the international standard AES-256. To meet the security compliance requirements of different countries and regions, it supports automatic switching of encryption standards based on the compliance requirements of the target country and generates encryption compliance reports that conform to local standards, ensuring legal and compliant operation globally.

[0339] Through the aforementioned system architecture, this application effectively addresses the challenges faced in implementing multilingual human-computer interaction methods across diverse devices and complex scenarios, including ensuring the immutability of core encoding rules, input process security, seamless module collaboration, and prevention of circumvention behaviors. The input module comprehensively and accurately captures multimodal user input across different devices, providing high-quality raw data for subsequent processing. The encoding module ensures that all input signals are uniquely and securely converted into internal codes through pre-stored, hardware-encrypted, and immutable number sequence vowel mapping rules. The built-in conflict detection unit further guarantees the validity of the encoding, fundamentally eliminating the risk of core encoding logic being tampered with or bypassed. The candidate processing module, based on an encrypted offline dictionary and combined with an intelligent sorting algorithm, provides users with a fast and accurate candidate list, greatly improving input efficiency and user experience. The display module ensures the interface's adaptability and visual consistency across various screen sizes, guaranteeing a cross-device user experience. Crucially, the control module adopts a multi-threaded collaborative architecture, achieving seamless linkage and mode switching between modules. Its built-in anti-circumvention monitoring unit is a core advantage of this system. This unit effectively identifies and intercepts any attempts to split or circumvent the core technical features of this method by rigorously verifying the module call chain, the timing continuity of input signals, and the integrity of core technical features, especially specific circumvention methods targeting small-screen devices. The deep integration of the hardware encryption unit and the encoding module provides physical-level security, ensuring the confidentiality and integrity of the core mapping rules and offline dictionary, and flexibly adapting to encryption compliance requirements worldwide. Therefore, this system not only implements the innovative input logic envisioned by this method, but also ensures the stability, reliability, and technical exclusivity of the method in practical applications through a rigorous and secure architecture, providing users with a secure, efficient, and consistent cross-device, multilingual human-computer interaction experience.

[0340] In the aforementioned multilingual human-computer interaction system based on the number-sequence vowel interface, although multiple input modes and interaction methods are provided, in actual use, users may frequently need to manually switch input modes to adapt to different input needs, such as switching from abbreviated pinyin to reduplicated pinyin, or manually activating tone mode when tone input is required. This manual switching operation not only increases the user's cognitive burden and operational steps, but may also lead to interruptions in the input process and reduced efficiency, especially in scenarios of rapid input or multilingual mixed input, resulting in a poor user experience.

[0341] To address this, this application further proposes an intelligent switching module that can automatically switch input modes based on user input sequences or multimodal interaction features. Specifically, when the system detects user input of "single / double-key initials," the intelligent switching module automatically switches the input mode to the stacked-key compatible abbreviated pinyin mode; when it detects user input of "initial + medial," the system automatically switches to the abbreviated pinyin extended mode; when it detects user input of "initial-medial-key double-key," the system automatically enters the stacked-key dedicated mode; when the user triggers tone input, the system automatically switches to the corresponding tone mode; when the user triggers multimodal input, the system automatically switches to the corresponding interaction mode; and when the system detects a single-key dual-phoneme encoding sequence, the system automatically switches to the corresponding mode. The delay of the entire switching process is controlled to ≤50ms.

[0342] The intelligent switching module is a functional module whose function is to automatically determine and switch to the most suitable input mode based on the user's input behavior or characteristics. This module can be a software or hardware module, responsible for real-time monitoring of the user's input actions and contextual information, and dynamically adjusting the system's input mode according to preset rules or machine learning models. Its implementation may include listening to input events (such as key events, voice events, and gesture events), analyzing the characteristics of input sequences (such as initial consonant combinations, medial vowel appearance, and tone trigger markers), and identifying the type of multimodal interaction (such as voice input and gesture input). Internally, this module contains a mode judgment logic unit and a mode switching execution unit. The mode judgment logic unit matches preset mode switching conditions based on input characteristics, while the mode switching execution unit is responsible for sending switching commands to the control module to activate the target input mode.

[0343] The automatic input mode switching based on user input sequences or multimodal interaction features aims to ensure that the system can automatically adjust to the most suitable input mode according to the user's actual input intention without manual intervention, thereby improving input efficiency and user experience. Specifically, the system continuously monitors the key sequence of user input. For example, when a continuous "single / double-key initial consonant" input pattern is detected, the system recognizes that the user may be performing superimposed input; when a combination of "initial consonant + medial consonant" is detected, the system determines that the user may need a simplified extended input mode. These sequence features are predefined and associated with specific input modes. Simultaneously, when the user inputs via voice, gestures, or other non-keyboard methods, the intelligent switching module analyzes the features of these multimodal input signals. For example, voice input may contain specific commands or language types, and gestures may correspond to specific functions or modes. The system will automatically switch to voice input mode, gesture control mode, or other corresponding interaction modes based on these features.

[0344] The system enters a consonant-based, abbreviated spelling mode when inputting single / double-key initials. This optimizes initial input processing, allowing the system to intelligently predict the user's intent and provide more flexible candidates. When a user inputs one or two keys that are recognized as initials (e.g., some keys are defined as initial keys according to the numerical vowel mapping rules), the intelligent switching module immediately switches the input mode to the consonant-based, abbreviated spelling mode. This means the system considers both consonant and abbreviated candidate words simultaneously, providing users with a wider range of choices.

[0345] The system switches to simplified pinyin expansion mode when the user inputs a "initial consonant + medial vowel". This is designed to accurately identify the user's intention to input the medial vowel and activate the simplified pinyin expansion mode to support more complex pinyin combinations. When the intelligent switching module detects that the user has inputted an initial consonant, followed immediately by a medial vowel (e.g., triggered by a long press or swipe), the system immediately switches to simplified pinyin expansion mode. In this mode, the system provides more accurate simplified pinyin expansion candidates based on the combination of the initial consonant and the medial vowel.

[0346] The input of "initial-medial-consonant double key" enters the stacking mode, which is designed to directly enter the stacking mode for specific initial-medial-consonant double key inputs, thereby improving input efficiency. When the user inputs two keys simultaneously or almost simultaneously, and these two keys are recognized as a combination of initial and medial consonants (e.g., through a specific double-key trigger mechanism), the intelligent switching module will directly switch the mode to the stacking mode, at which time the system will mainly provide stacking-related candidates.

[0347] The feature of switching to the corresponding tone mode upon triggering tone input aims to ensure that tone input can be seamlessly integrated into the overall input process, eliminating the need for users to manually select a tone mode. When a user triggers tone input using a single tone (e.g., numbers 1-4), a double tone (e.g., numbers 6, 7-0), or a character tone (e.g., long-pressing a punctuation mark), the intelligent switching module immediately recognizes this operation and switches the system to the corresponding tone mode. At this time, the candidate processing module generates a list of candidate syllables with tones based on the input initial-medial-final combination and tone information.

[0348] The feature of switching to the corresponding interaction mode upon triggering multimodal input aims to enable the system to flexibly respond to various input methods such as voice and gestures, providing a seamless multimodal interactive experience. When the intelligent switching module detects voice input (e.g., via voice wake-up or voice-to-text function), gesture input (e.g., via screen gesture recognition), or other multimodal input events, the system will immediately switch to the corresponding interaction mode. For example, during voice input, the system will activate speech recognition and text transcription functions; during gesture input, the system will activate gesture recognition and corresponding function execution.

[0349] The automatic switching of the corresponding mode upon detecting a single-key dual-phoneme encoding sequence aims to intelligently identify and switch to a mode supporting this function for special input methods involving single-key dual-phonemes, ensuring correct encoding parsing and candidate generation. When the intelligent switching module detects that the user has triggered a single-key dual-phoneme combination by long-pressing the 0 key or the "-" key, and generated the corresponding encoding sequence, the system will automatically switch to the mode supporting single-key dual-phonemes. In this mode, the encoding module will correctly parse the encoding sequence, and the candidate processing module will generate abbreviated codes or syllable symbols for initial-medial-vowel-tone syllables containing medial vowels.

[0350] The switching latency of ≤50ms aims to ensure smooth and real-time mode switching, avoiding noticeable delays perceived by users and thus improving user experience. This requires the control module to adopt an efficient multi-threaded collaborative architecture, optimize the mode judgment logic and switching execution process, and reduce unnecessary computation and resource consumption. For example, resources for commonly used modes can be preloaded, or a lightweight mode switching mechanism can be adopted to ensure that mode recognition and switching are completed within 50 milliseconds.

[0351] By introducing an intelligent switching module, this application solves the problem of cumbersome operation and reduced efficiency caused by users having to manually switch modes in multi-modal input scenarios. This module can intelligently and seamlessly switch to the most suitable input mode automatically based on the user's real-time input sequence and multimodal interaction characteristics. For example, it automatically enters the overlapping and abbreviated pinyin mode when inputting initials, automatically switches to tone mode when triggering tones, and automatically switches to the corresponding mode when a single-key dual-phoneme encoding sequence is detected. This automated switching mechanism greatly simplifies the user's operation process, eliminates the cognitive burden and time consumption of manual switching, and allows users to input in a more natural and fluent way. At the same time, a switching latency of ≤50ms ensures the real-time and seamless nature of mode switching, avoids input interruption, and significantly improves overall input efficiency and user experience. Combined with the encoding module, candidate processing module, and control module in the above system, the intelligent switching module makes the entire interactive system more intelligent and user-friendly, especially in scenarios such as fast input, multilingual mixed input, and small-screen devices, where its advantages are more obvious, effectively improving the fluency and accuracy of human-computer interaction.

[0352] In some of the embodiments described above in this application, although the system can efficiently process user input and generate candidate lists, there is still room for improvement in handling complex semantic understanding, contextual relationships, and user intent prediction. For example, in long text input or multi-turn dialogue scenarios, relying solely on a preset lexicon and frequency ranking may not provide the most accurate and context-appropriate candidate words or phrases, nor can it intelligently recommend specific word combinations with tones, thereby affecting user input efficiency and experience.

[0353] To address this, this application further proposes a human-computer interaction method, which also includes an AI collaboration module. This AI collaboration module aims to interact with a generative AI model to achieve semantic-level input completion and multi-turn dialogue encoding collaboration based on numerical vowel encoding logic. It also recommends high-frequency single-key diphone combinations based on the input context to achieve tone-based vocabulary completion.

[0354] Specifically, the AI ​​collaboration module is an intelligent processing unit within the system. Its core function is to leverage advanced artificial intelligence technology to enhance the intelligence level of the input system. This module surpasses traditional candidate generation methods based on word frequency or simple semantic association, gaining a deeper understanding of the user's input intent and contextual information. The AI ​​collaboration module can be deployed as a standalone software component or integrated into the system's core processing logic, responsible for managing communication with external or built-in AI models, data preprocessing, and result parsing.

[0355] The AI ​​collaboration module can interact with large generative AI models. This interaction is typically achieved through a standardized application programming interface (API), allowing the system to send partial user input, current dialogue context, and other information to the remotely or locally deployed generative AI model. Leveraging their massive training data and complex neural network structures, these generative AI models can understand and generate high-quality natural language text, thus providing the input system with deeper semantic analysis and predictive capabilities. During the interaction, the system ensures the security and efficiency of data transmission, for example, by employing encrypted communication protocols and compressing data to reduce latency.

[0356] Based on the numerical vowel encoding logic, the AI ​​collaborative module can achieve semantic-level input completion. This means that the system can not only predict the next possible character or word based on the numerical vowel encoding sequence already entered by the user, but also understand the overall semantics of the user's input and complete the entire phrase or sentence accordingly. For example, when a user enters "wo xiang yao ding yi zhang", the system can intelligently complete it as "wo xiang yao ding yi zhang ji piao" (I want to book a plane ticket) based on the context, instead of just providing the completion of a single character. This semantic-level completion capability greatly improves the efficiency of users inputting long texts or complex sentences.

[0357] Furthermore, the AI ​​collaboration module also features multi-turn dialogue encoding collaboration. In multi-turn dialogue scenarios, the system can remember and maintain the context information of the dialogue, using previous communication content as an important reference for the current input. When the user makes subsequent inputs, the AI ​​collaboration module submits the current numerical vowel encoding sequence along with the historical dialogue context to a generative AI model for analysis, thereby generating candidate words or phrases highly relevant to the entire dialogue process. For example, after asking about the weather, if the user enters a location again, the system can automatically connect it to the context of the weather query and provide more accurate location-related suggestions. This collaborative function ensures the coherence and intelligence of the input process, making human-computer interaction more natural and fluid.

[0358] To ensure the reliability of the above functions, the AI ​​collaboration accuracy can reach 92% or higher. This high accuracy is achieved through continuous optimization of the AI ​​model, expansion of the training dataset, and the adoption of advanced algorithms. The system regularly evaluates and calibrates the performance of the AI ​​collaboration module to adapt to language changes and the evolution of user habits.

[0359] Furthermore, the AI ​​collaboration module can recommend high-frequency single-key diphone combinations based on the input context. During user input, the AI ​​collaboration module analyzes the current numerical vowel encoding sequence and its context in real time, intelligently determining which single-key diphone combinations (e.g., combinations of medial vowels and tones) appear more frequently or have strong semantic relevance in the context. For example, when a user inputs a combination of an initial consonant and a final vowel, the system predicts the most likely medial vowel and tone combination based on the context and recommends it to the user as a high-priority candidate, thereby simplifying the input process of complex syllables.

[0360] Through the aforementioned technical solution, this application overcomes the limitations of traditional input methods in semantic understanding and contextual association. Based on numerical vowel encoding logic, this module achieves semantic-level input completion, enabling the system not only to predict individual words but also to intelligently complete entire phrases or sentences according to context, significantly improving input efficiency and accuracy. Simultaneously, through multi-turn dialogue encoding collaboration, the system can memorize and utilize dialogue history, ensuring coherent and context-appropriate candidates in continuous communication. Furthermore, the recommendation of high-frequency single-key diphone combinations and tone-based vocabulary completion based on input context provides users with more accurate and convenient suggestions when inputting complex, tone-based language, effectively reducing the input error rate and significantly optimizing the user's input experience in complex language environments.

[0361] In some of the embodiments described above in this application, a multilingual human-computer interaction system is proposed. This system achieves efficient and secure input based on a number-sequential vowel interface through the collaborative work of an input module, an encoding module, a candidate processing module, a display module, a control module, and a hardware encryption unit. However, in practical applications, this system may face risks such as cross-operating system compatibility, compliance with data security regulations in different countries, and the risk of its core technical features being counterfeited or circumvented. These issues may limit the widespread deployment of the system and the effective protection of intellectual property rights.

[0362] To address this, this application further proposes a multilingual human-computer interaction system that supports Android 8.0+, iOS 12.0+, Windows 10+, Linux (Ubuntu 18.04+), and industrial real-time operating systems (VxWorks 7.0+, QNX 7.0+). The system boasts a cross-platform compatibility rate of at least 99%, with a basic installation package size of no more than 20MB and a full version size of no more than 50MB. To meet the patent examination requirements of different countries, the system supports compliant adaptation of configuration parameters for encoding modules and encryption units, satisfying data security and encryption standards of various countries without altering the core claims. Furthermore, the system features a patent infringement warning function, capable of monitoring and recording counterfeit use of core encoding rules and parent-child interface logic, and generating an infringement evidence chain containing timestamps and device identifiers. Furthermore, the system adds a technical feature splitting and circumvention verification scheme, specifically including: test scenarios, simulating the splitting of the "three-in-one mapping table" as described in claim 1 (only implementing vowel-key mapping, skipping encoding sequence binding), the splitting of the "single-key dual-phoneme triggering logic" as described in claim 21 (first triggering the key and then generating dual-phoneme encoding through a third-party module), and the splitting of the "module collaboration logic" as described in claim 27 (only calling some modules to achieve equivalent input); test methods, under a standard test environment (room temperature 25℃±5℃, humidity 40%-60%, device memory ≥2GB), generating the above splitting and circumvention input behaviors through automated testing tools, and recording the system response results; judgment criteria, if the system can 100% identify and intercept the above splitting and circumvention behaviors, and generate a warning log containing "split behavior type, trigger time, and device identifier", it is judged as passing the verification; if there are unintercepted splitting and circumvention behaviors, it is judged that the core technical features are not effectively protected, and anti-circumvention logic optimization needs to be initiated.

[0363] The aforementioned system significantly expands its application scope and user base by supporting multiple mainstream operating systems. To achieve broad cross-platform compatibility, the system can be developed using cross-platform frameworks (such as Qt and Flutter), or provide software development kits (SDKs) or application programming interfaces (APIs) for different operating systems. This ensures that the underlying core logic remains consistent across platforms, while the user interface and system interaction layer are adapted to the specific characteristics of each operating system. High compatibility means the system can provide stable and consistent functionality and user experience across different operating systems. Furthermore, through techniques such as code streamlining, modular design, on-demand module loading, and optimized resource files, the system achieves a smaller installation package size, thereby reducing download and storage costs for users and improving deployment efficiency.

[0364] To meet the patent examination requirements and data security standards of different countries, the system supports compliance adaptation of the configuration parameters of the encoding module and hardware encryption unit. This can be achieved by pre-setting or dynamically loading compliance configuration files for different regions within the system. For example, in terms of data encryption, the system can dynamically select encryption algorithms such as the Chinese national standard SM4 or the international standard AES-256 according to the regulatory requirements of the target country, and adjust parameters such as key length and encryption mode. This adaptation mechanism ensures the legal and compliant deployment of the system globally, effectively avoiding legal risks caused by regional differences, while not affecting the integrity of core encoding rules and mapping relationships.

[0365] Furthermore, the system integrates a patent infringement early warning function, designed to proactively monitor and record counterfeit use of core encoding rules and parent-child interface logic. This function is implemented by embedding a monitoring unit within the control module, which continuously analyzes the system's internal input processing flow, encoding generation logic, and parent-child interface interaction patterns. When it detects behavior that does not conform to the core technical features of this method but attempts to achieve equivalent functionality—for example, modifying vowel mapping relationships without authorization, bypassing hardware encryption units to access the core dictionary, or calling modules in a non-cooperative manner—the system will trigger an early warning. The warning information will record detailed key data such as the type of behavior, the time of occurrence, and the identification of the devices involved, and generate an immutable chain of evidence, providing strong support for subsequent intellectual property protection.

[0366] To further strengthen the protection of core technical features, the system has added a technical feature splitting and circumvention verification scheme. This scheme tests the system's anti-circumvention capabilities by simulating specific splitting behaviors. For example, in the test scenario, the system will simulate behavior that only implements the mapping between vowels and keys, skipping the binding of the encoding sequence; or simulate behavior that receives key triggers first and then generates dual-phoneme codes through an external module to circumvent the built-in trigger logic; and simulate equivalent input behavior that only calls some modules (such as the input module and the display module) to bypass the encryption verification and collaborative processing of the encoding module. In a standard test environment, these simulated splitting and circumvention input behaviors are generated using automated testing tools, and the system's response results are recorded. The judgment criterion is that the system must be able to 100% identify and intercept these unauthorized or splitting behaviors and generate a warning log containing the splitting behavior type, trigger time, and device identifier. If the system fails to completely intercept, it indicates that there is a vulnerability in the anti-circumvention logic, and an optimization process needs to be initiated to enhance its robustness.

[0367] Through the aforementioned technical solutions, this system significantly expands its application scope and user base, ensuring stable operation and a consistent user experience across different hardware and software environments, and avoiding technical fragmentation and deployment obstacles caused by platform limitations. Simultaneously, by adapting the encoding module and encryption unit for compliance, this system can flexibly meet data security and encryption standards worldwide, enabling legal and compliant deployment and promotion globally, effectively mitigating legal risks associated with international market access. More importantly, by integrating patent infringement warning functions and technical feature decomposition and circumvention verification schemes, this system can proactively monitor and effectively prevent the imitation and use of core encoding rules and parent-child interface logic, as well as technical decomposition and circumvention behaviors. This not only provides strong intellectual property protection for the innovative technology of this method, ensuring that core technological advantages cannot be easily copied or circumvented, but also guarantees the effectiveness and robustness of the system's anti-circumvention logic through a continuous verification mechanism, thereby maintaining the market competitiveness and uniqueness of this technical solution.

[0368] The following example will provide a more detailed explanation of the above technical solution:

[0369] In a typical multi-device usage scenario, User A wants to seamlessly switch between a smartwatch and a mobile terminal, perform mixed Chinese and English input, and ensure input efficiency and accuracy.

[0370] First, this method defines a unique set of numerical vowels and mapping rules. The system pre-defines 11 core vowels (ɑ, ɑnɡ, engɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi) in various keyboard formats as shown in Figures 1-4, 42, and 44. These vowels are formed by splitting or merging according to the principle of "consistency of tongue position in pronunciation" for medial vowels in Pinyin. For example, the vowel "ɑ" is mapped to the numeric key "1", the vowel "ɑnɡ" is mapped to the numeric key "2", and so on. This three-in-one mapping table of "vowel-key identifier-encoding sequence" is the core of the system, stored through a hardware chip-level encryption unit, and cannot be modified through conventional software configuration, ensuring the stability and security of the encoding. This is fundamentally different from existing input methods that rely on alphabetic keyboard character mapping; this method constructs an independent numerical vowel encoding system. After obtaining intellectual property protection, namely PCT invention patents and patent authorizations from relevant countries, character mapping input can be used to store the data through software encryption units, thus constructing an independent numerical vowel encoding system.

[0371] Next, the system is configured with a cross-device compatible digital keypad and extended keypad system. On User A's smartwatch, due to the small screen size, the system automatically enables the extended keypad identifiers "11" and "12". For example, the "11" key is configured to trigger single-key dual-phoneme combinations related to the medial vowel "i", while the "12" key is used to trigger combinations related to the medial vowels "u / ü". These extended keypad identifiers, together with the core digital keypads (1-0), are incorporated into the aforementioned three-in-one mapping table, and their mapping relationship cannot be modified or separated individually. When User A switches from the smartwatch to the mobile terminal, the two devices automatically achieve seamless synchronization of the number sequence vowel mapping rules and candidate prompt sequences via the Bluetooth HID protocol, with a synchronization latency of less than 50ms. This means that User A's input logic and candidate word order, which are accustomed to on the smartwatch, remain completely consistent on the mobile terminal, avoiding the problem of fragmented input logic across devices in existing technologies.

[0372] On this basis, the system constructs an irreplaceable numerical sequence-vowel interface. On a smart watch, the interface contains 10-16 keys, and the key layout follows the "minimal layout principle for small-screen devices". The "11" and "12" extended keys are preferentially used to achieve single-key triggering of core vowels. For example, long-pressing the "11" key can directly input the "i" medial, and then combined with the numeric keys to input the vowel. On a mobile terminal, the interface layout follows the "principle that high-frequency vowels are close to the hand", binding high-frequency vowels (such as ɑ, ɑn, en) to keys 1-3 or keys 1, 4, 5 to improve input efficiency. Regardless of which device, the interface layout is strongly bound to the numerical sequence-vowel mapping rule, and the correspondence between the core keys and vowels cannot be modified through the conventional interface customization function. The key layout on the smart watch will be synchronized to the mother-child interface system of the mobile terminal, ensuring the consistency of the input logic across devices.

[0373] Subsequently, the system constructs a coordinated mother-child interface system. User A inputs the numerical sequence-vowel encoding of "ni hao" on the smart watch, and the system displays the candidate word "你好". At this time, User A picks up the mobile terminal, and the system realizes the seamless switching of the mother / child interface through the control module (the switching delay is less than 30ms). The interface of the mobile terminal immediately displays the same "你好" candidate list, and the context of "ni hao" that User A has input on the watch is maintained without interruption. This is because the mother-child interface shares the same encoding mapping table and candidate word library, ensuring that the same candidate list is output when the same initial consonant-medial combination is input, solving the problem of inconsistent candidate sequences and low coordination efficiency between different forms of keyboards in the prior art. It can also adopt the same numerical sequence-vowels and the same columnar character layout as the alphabetic keyboard, as shown in Figure 44, to form a coordinated mother-child interface system of multiple keyboards.

[0374] Furthermore, the method realizes a unified encoding logic for multiple languages. User A needs to input a text containing a mixture of Chinese and English, such as "Hello你好". Based on the core encoding rule of numerical sequence-vowels, the system realizes the unified encoding of Chinese, English, and other tonal languages through "vowel reserved encoding positions + voice feature mapping". User A can directly input "Hello ni hao" without switching the language mode, and the system can recognize and generate the corresponding Chinese-English mixed candidates. There is no conflict between the encoding sequences of different languages, greatly improving the efficiency of mixed multi-language input and overcoming the disadvantages of independent encoding systems for each language, overlapping character keys, and logical conflicts during bilingual mixed input in the prior art.

[0375] The technical features 1)-5) mentioned above constitute an inseparable whole for realizing its core function and must be implemented in a coordinated manner. For example, if a scheme attempts to implement only the vowel-key mapping but skips the binding of the encoding sequence, or splits the triggering operation of single-key dual-phoneme and the encoding generation into independent steps, the anti-avoidance monitoring unit of this method will identify and intercept such splitting and circumventing behavior by detecting the temporal continuity of the input operation (the core operation interval is less than 100ms) and the integrity of the encoding sequence (whether any link in the three-in-one mapping table is missing). This emphasizes that the numerical vowel encoding system of this method is a unified core framework. All cross-device adaptation, scenario expansion, and mode optimization must be implemented based on this framework. Independent input logic cannot be built outside the framework, thereby effectively distinguishing it from existing input method technologies based on letter key reuse and preventing the technical features from being split and circumvented.

[0376] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

[0377] This technical solution fully meets the industrial applicability requirements of PCT applications and can be manufactured, integrated, and applied on a large scale in multiple industrial sectors worldwide, generating significant economic and social value.

[0378] 1. Device compatibility: It can be seamlessly integrated into all types of electronic devices such as mobile terminals, smart wearables, industrial control equipment, IoT devices, vehicle devices, VR / AR devices, medical terminals, and cross-border communication devices. It is compatible with Android, iOS, Windows, Linux and industrial real-time operating systems, and is compatible with full-size screens from 1.5 to 21 inches. The hardware requirements are in line with the configuration of conventional electronic devices, without the need for special customized hardware, and it has industrial feasibility for mass integration.

[0379] 2. Industry Applications: Covering multiple core industries including children's education, industrial control, cross-border communication, medical data entry, financial office work, autonomous driving, cross-border e-commerce, and remote collaboration. In children's education, it can serve as an auxiliary tool for pinyin teaching; in the industrial sector, it can adapt to core command input in offline environments; in cross-border scenarios, it can solve the need for multilingual offline communication; and in the medical and financial sectors, it can improve the efficiency of inputting professional terminology with tones. Each industry application addresses practical pain points and possesses clear industrial application value.

[0380] 3. Technical implementation: The core coding rules, collaborative logic, and anti-circumvention mechanisms all have clear software implementation paths. The offline dictionary, encryption algorithm, and multimodal interaction module can be integrated using conventional software development tools. Test indicators (such as response latency, recognition accuracy, and compatibility) can all pass industrial-grade testing and verification, meeting the quality control requirements in large-scale production.

[0381] 4. Commercial Value: It solves the industry pain points of traditional input methods, such as fragmented cross-device adaptation, inefficient voice input with tone, insufficient multi-scenario adaptation, and easy circumvention. The product can be launched in the global market as an independent input method application, embedded input module, or industry-customized solution. It adapts to the language habits and compliance requirements of different countries and has broad commercial promotion prospects and sustainable profit model.

[0382] Table 1: Examples of phonetic characters used for elementary learning with the fewest characters and their phonetic symbols

[0383] Table 1: Examples of phonetic characters for beginners with the fewest characters and their phonetic symbols (continued 1)

[0384] Table 1: Examples of phonetic characters for beginners with the fewest characters and their initial and final symbols (continued 2)

[0385] Table 1: Examples of phonetic characters for beginners with the fewest characters and their phonetic symbols (continued 3)

[0386] Table 2: Syllable Codes for Initials, Medials, Finals, and Supplementary Numbers in Alliteration

[0387] Table 2: Syllable abbreviations for alliterative words with initial consonant, medial vowel, and additional numerical order (continued 1)

[0388] Table 2: Syllable abbreviations for alliterative words with initial consonant, medial vowel, and additional numerical order (continued 2)

[0389] Table 2: Syllable abbreviations for alliterative words with initial consonant, medial vowel, and additional numerical order (continued 3)

[0390] Table 3: Abbreviated Codes for Numbered Initials and Tones of Finals with Additional Numbers

[0391] Table 3: Abbreviated Codes for Numbered Initials and Tones of Finals with Additional Numbers (Continued 1)

[0392] Table 4: Abbreviated Codes of Numbered Syllables for Initials and Tones with Additional Numbered Finals

[0393] Table 4: Abbreviated Codes of Numbered Syllables for Initials and Tones with Additional Numbered Finals (Continued 1)

[0394] Table 5: A schematic diagram of the graphical user interface for the phonetic-character-word-sentence human-computer interaction method in the teaching interface.

Claims

1. A multilingual human-computer interaction method based on a number-sequence vowel interface, characterized in that, include: 1) Define a unique set of numerically ordered finals and mapping rules: The numerically ordered finals are formed by splitting or merging the medial finals of Chinese Pinyin according to the principle of "consistency of tongue position in pronunciation", with a total of 11 groups of core finals (ɑ, ɑnɡ, engɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi). Each group of finals corresponds to a unique digital key position, and a unique mapping table of "final - key position identifier - encoding sequence" is established. The mapping table cannot be modified by conventional software configuration. 2) Configure a cross-device compatible numeric keypad and extended keypad system: Extended keypad identifiers are selected from "11" and "12", and extended characters are selected from one or more of "-", "=", "*", and "#". Among them, "11" and "12" are specifically used to adapt to the keypad extension of small-screen smart wearable devices such as smartwatches. The extended keypad identifiers and core numeric keypads (1-0) are included in the "vowel-keypad identifier-encoding sequence" three-in-one mapping table, and their mapping relationship cannot be modified or split separately. Different devices (mobile terminals, smart wearables, industrial control equipment) share the same numeric vowel mapping rules and candidate prompt sequences, and the mapping rules are seamlessly synchronized through USB HID protocol and Bluetooth HID protocol, with a synchronization delay of ≤50ms. 3) Construct an irreplaceable numerical vowel interface: The interface contains 10-16 keys (including "11" and "12" extended keys). The key layout follows the "high-frequency vowels close to the hand principle" and the "minimalist layout principle for small screen devices". Small screen devices such as smartwatches prioritize the use of "11" and "12" extended keys to expand the core vowels and achieve single-key triggering of vowels. Furthermore, the interface layout is strongly bound to the numerical vowel mapping rules, and the correspondence between the core keys and vowels cannot be modified through the conventional interface customization function. The key layout of small screen devices such as smartwatches needs to be synchronized with the parent-child interface system to ensure the consistency of input logic across devices. 4) Construct a collaborative parent-child interface system: The keyboard and its extended interface with the same number sequence vowels and consistent number sequence string layout and sorting constitute the parent-child interface. They share the same encoding mapping table and candidate word library. When the same initial consonant combination is input, the same candidate list is output. The control module realizes the zero-delay switching of the parent / child interface (switching delay ≤30ms) and keeps the input context uninterrupted during the switching process. 5) Multilingual unified encoding logic: Based on the core encoding rules of the numerical vowels, unified encoding of Chinese, English and other languages ​​with tones is achieved through "vowel reserved encoding position + speech feature mapping". Bilingual mixed input can be completed without switching language modes, and there is no conflict between the encoding sequences of different languages. 6) Core Technology Differentiation and Indivisible Features: The method does not rely on the character mapping relationship of the existing letter keyboard, but achieves input through an independent numerical vowel encoding system, which is different from the existing input method technology based on letter key reuse; and the aforementioned technical features 1)-5) are an indivisible whole for realizing the core function of this method, which needs to be implemented in a coordinated manner. Any scheme that breaks it down into independent steps or modules and only implements some features falls within the protection scope of this method; furthermore, the identification standard and technical logic for the splitting of core technical features are as follows: if a scheme breaks down "vowel mapping - key configuration - interface construction - collaborative interaction - encoding" into an indivisible whole, it will be considered as follows: Any attempt to bypass the core mapping table or collaborative logic to achieve equivalent input functionality by modifying the core mapping relationship or splitting the association between extended keys and core keys under the guise of "device-specific customization" is considered a splitting circumvention. The technical identification of splitting circumvention is achieved by detecting the temporal continuity of input operations (core operation interval ≤ 100ms) and the completeness of the encoding sequence (whether any link in the three-in-one mapping table is missing). The number sequence vowel encoding system of this method is a unified core framework; all cross-device adaptations, scenario expansions, and mode optimizations must be implemented based on this framework, and independent input logic cannot be built outside of it.

2. The method as described in claim 1, characterized in that, The specific configuration of the number sequence vowel interface includes any of the following types: 1) 11-key basic vowel interface: It consists of the keys corresponding to 1-9, 0 (corresponding to 10) and the extended character "-" (marked as 11); 2) 10-key compressed vowel interface: Combines the vowel combinations corresponding to two adjacent keys in the 11-key interface, consisting of keys 1-9 and 0; 3) 12-key multilingual vowel interface: Based on the 11-key interface, a new key (marked as 12) corresponding to the extended character "=" is added, reserving space for multilingual encoding; 4) 15-key multilingual vowel interface: Add 4 character tone keys T1-T4 to the 11-key interface, or use a shared key layout for "punctuation-tone"; 5) 16-key full-function character layout: Add 4 tone keys T1-T4 to the 12-key layout, or use a shared key layout for "punctuation-tone".

3. The method as described in claim 1, characterized in that, The specific construction method of the set of numbered vowels is as follows: the medial vowels other than uo and üe are split into medial vowels + vowels and merged into 11 sets of adaptable vowels (ɑ, ɑnɡ, enɡ, ɑn, en, e / üe / er, u / o / uo / i, ɑo, ei / ü, ou, ɑi), where the positions of ɑ and u / o / uo / i are fixed, en is only interchanged with e / üe / er, and the other vowels can be interchanged arbitrarily; when there is no need for adaptation sharing, the position of each vowel is not restricted; two sets of vowels can be merged into 10 sets of compressed numbered vowels.

4. The method as described in claim 1, characterized in that, The 11 sets of numerical vowels can be sorted using any of the following schemes: (1) Efficient input scheme: high-frequency vowels (ɑ, ɑn, en) are bound to keys 1-3, and low-frequency vowels (ei / ü, ou, ɑi) are bound to keys 9, 0, 11; (2) Alphabet keyboard compatible scheme: sorted according to the order of the numeric keys on the alphabetical keyboard; (3) Children's teaching scheme: sorted according to pronunciation difficulty (simple vowels → compound vowels → nasal vowels).

5. The method as described in claim 1, characterized in that, The dynamic keyboard supports layered display: the basic layer (small screen ≤ 3.5 inches) retains only keys 1-12 (including extended keys 11 and 12), occupying ≤ 30% of the screen height; the standard layer (medium screen 3.5-7 inches) includes all keys + tone keys + mode switching keys, occupying 40%-50% of the screen height; the extended layer (large screen ≥ 7 inches) supports shortcut key binding + reserved labels + single key dual-phoneme candidate area, occupying 50%-60% of the screen height.

6. The method as described in claim 1, characterized in that, The dynamic keyboard and external hardware keyboard are automatically detected via USB HID and Bluetooth HID protocols, with a detection latency of ≤100ms. It supports automatic hiding / synchronous display / floating mode switching with a switching latency of ≤50ms. When the hardware keyboard inputs a single key with a dual-phoneme combination, the corresponding key on the dynamic keyboard is highlighted synchronously. For small-screen devices such as smartwatches, touch optimizations have been added for the "11" and "12" extended keys: support adaptive touch area (minimum touch area ≥8×8mm), accidental touch filtering (the pressure threshold can be customized, ranging from 5-20g), and touch recognition accuracy ≥98% (test environment: room temperature 25℃±5℃, humidity 40%-60%, small-screen device screen size ≤2 inches, processor ≥Snapdragon 855 / Apple A13, memory ≥2GB).

7. A method for extending the tone of a number sequence based on the method described in claim 1, characterized in that, include: 1) Single tone function: After inputting the initial-medial-final combination, the numbers 1-4 correspond to the first-fourth tones of Chinese, and the number 5 corresponds to the neutral tone or no tone. The candidate list of syllables with tone is displayed simultaneously. 2) Number sequence dual tone function: Add a set of number sequence tone configurations. The number 6 corresponds to the neutral tone or no tone, and the numbers 7-0 correspond to the fourth tone and the first to third tone of Mandarin, respectively. It supports custom sorting and forms a dual tone system with the number sequence single tone. It can simultaneously prompt two sets of candidate syllables with tone. 3) Candidate prompt rules: The extended interface determines the maximum number of prompt strings per page based on the number sequence vowel, key position identifier and extended key position identifier, number sequence string, multilingual characters and screen size.

8. The method as described in claim 7, characterized in that, The custom sorting of the numerical dual tones supports users setting a "high-frequency tone priority" rule: in adult mode, the four tones of Mandarin (usage rate ≥35%) are bound to key 7, and in children's mode, the first tone and neutral tone (total usage rate ≥40%) are bound to key 6-7. Custom rules can be synchronized to the parent-child interface, and the tone input error rate is ≤3%.

9. The method as described in claim 7, characterized in that, The number of string prompts on each page of the extended interface is dynamically adjusted according to the screen size: (1) Small screen devices (screen size ≤ 2 inches, including smartwatches): 2-3, character size ≥ 16, supporting up and down swiping to switch pages; (2) Small screen devices (screen size 2-3.5 inches): 3-5, character size ≥ 14; (3) Medium screen devices (3.5-7 inches): 6-8, character size ≥ 12; (4) Large screen devices (≥ 7 inches): 9-11, supporting left and right swiping to switch pages; The candidate prompt rules for all devices are strongly bound to the number sequence vowel mapping rules and cannot be modified independently.

10. A dual-mode tone interface human-computer interaction method based on the method of claim 1, characterized in that, include: 1) Character tone configuration: Includes independent tone keys and shared keys. The size of the independent keys is the same as that of the numeric keys, which can be adapted to a 4-row 4-column or circular layout; the shared key mapping relationship is ",-T1, .-T2, ?-T3, !-T4" and / or ",-T1, .-T2, / -T3, ]-T4 or \-T4"; 2) Seamless expansion of consonant set: No need to switch input modes, supports two input rules - ① initial consonant + medial vowel + character tone + final vowel; ② initial consonant + character tone + medial vowel + final vowel, sharing the number sequence final vowel mapping and candidate logic with the main consonant set; 3) Dual-mode collaboration: Number tone and character tone can be switched freely, and a tone combination vocabulary mapping library is shared. When the same tone identifier is input, the candidate results are consistent. When there is a conflict, priority is given to the identification and the operation is triggered first. The system automatically remembers the most recently used tone mode.

11. The method as described in claim 10, characterized in that, The triggering logic for the shared "punctuation-tone" key is as follows: a light touch (press duration ≤ 200ms) outputs a punctuation mark, a long press (press duration ≥ 300ms, ≥ 400ms in child mode) pops up a tone candidate list (T1-T4), and the tone is output by sliding to the corresponding tone key and releasing it during the long press; the user can customize the duration threshold of the light touch / long press (range 100-500ms).

12. The method as described in claim 10, characterized in that, The consonant set expansion is automatically activated in the following scenarios: (1) If no further input is received within 500ms after inputting "initial consonant + medial consonant", the consonant set candidate will automatically pop up; (2) In multilingual mixed input scenarios, automatic parsing is performed according to consonant set rules; (3) In children's teaching scenarios, the consonant set candidates are displayed in a synchronous animation of the pinyin splitting of "initial consonant + medial consonant + tone".

13. A generalized dynamic extension interface human-computer interaction method based on the method described in claim 1, characterized in that, include: 1) Flexible wake-up methods: Supports shortcut keys, voice commands, gestures, floating icons, or user-defined trigger methods. On Android 8.0 and iOS 12.0 and above mobile terminals, the trigger response latency is ≤80ms (test environment: room temperature 25℃, no external interference); the voice wake-up word can be customized, and the recognition accuracy is ≥92% in ambient noise below 50dB and in the frequency range of 200-3000Hz. Offline wake-up is supported. 2) Core interface configuration: Includes associated prompt area, dynamic candidate area, fault-tolerant prompt area, and multimodal interaction area. The multimodal interaction area is compatible with the input logic mapping of hard keyboard, static soft keyboard, and dynamic soft keyboard. 3) Multimodal interaction extension: Supports multimodal input and output such as speech-to-text, text-to-speech, gesture control, and image character recognition. Multimodal data needs to be mapped to the corresponding numerical vowel codes, and candidates are generated in conjunction with key input codes and synchronized to the dynamic candidate area in real time. Among them, speech-to-text supports ≥8 common languages, with an online recognition accuracy of ≥95% and an offline recognition accuracy of ≥90%. Image character recognition supports multiple languages ​​with an accuracy of ≥90%. 4) Adaptation and optimization: The interface can be dragged and scaled (range 50%-200%), the font size of the dynamic candidate area is adjustable from 8 to 24 points, and it supports switching between dark / light / high contrast modes; the dynamic candidate area and the trigger operation area are horizontally aligned, with an alignment error of ≤1 pixel.

14. The method as described in claim 13, characterized in that, The fault-tolerant prompt area calculates the probability of mis-touch / mis-recognition based on the physical distance of the key position or the similarity of voice recognition. The accuracy of the Top 1 correction candidate is ≥92%, and the correction logic is consistent with the number sequence vowel mapping.

15. The method as described in claim 13, characterized in that, Multimodal interaction area mode adaptation: (1) Adult mode: enhanced professional terminology recognition, accuracy ≥90%; (2) Children mode: enhanced pronunciation error tolerance, non-standard pronunciation recognition accuracy ≥88%, picture book character recognition accuracy ≥90%; (3) Single key dual phoneme collaboration: after speech recognition with syllable adjustment, the corresponding key on the dynamic keyboard is highlighted synchronously.

16. A method for scene adaptation, multi-directional layout adaptation, and core engine-supported human-computer interaction based on the method described in claim 1, characterized in that, include: 1) Cross-device adaptation: Configure the effective touch area of ​​the key surface according to the device type: mobile terminal ≥8×8mm, smart wearable device (including smartwatch) ≥10×10mm ("11" and "12" extended keys are the same as standard), commercial device ≥12×12mm, industrial control panel ≥15×15mm; the core input logic and number sequence vowel mapping rules of all devices remain completely consistent. Only the physical size and layout density of the keys are allowed to be adjusted. The core coding association relationship cannot be modified. 2) Multi-directional layout adaptation: Supports right-angle left vertical, right-angle right vertical, and top reverse arrangement. The character direction of the top reverse arrangement can be selected by the user. It supports manual switching or automatic adaptation according to the document layout direction, and the switching delay is ≤50ms. 3) Scene customization and adaptation: Children's teaching scenarios are equipped with pronunciation and mouth shape animation display, 3-level speech speed adjustment and follow-up feedback function; industrial control scenarios optimize core command input and caching; Enhanced offline multilingual translation and pronunciation prompts based on numerical vowel encoding for cross-border communication scenarios; 4) Multilingual extension and adaptation: Based on speech features, core characters of other languages ​​are bound to reserved encoding positions of number vowels, supporting standard pronunciation prompts for words in multiple languages ​​(naturalness ≥ 4.0 points); 5) Core Support Engine: Integrates offline encoding verification based on number-sequential vowel encoding, mixed-mode parsing, offline dictionary indexing, and multilingual bidirectional offline translation functions. The offline dictionary contains ≥80,000 entries for Chinese, ≥50,000 entries for English, and ≥30,000 entries / language for other languages.

17. The method as described in claim 16, characterized in that, The visual teaching functions of the children's teaching scenario include: animated display of the pronunciation mouth shape of Chinese Pinyin, 3-level speech speed adjustment (50 / 100 / 150 words / minute), interactive syllable reading game, encouraging feedback after correct input, and visual split display of single-key double phoneme combination.

18. The method as described in claim 16, characterized in that, The offline encoding verification module of the core support engine rejects invalid combinations such as "zh+ü" and "b+iong", and the verification response delay is ≤150ms (test environment: room temperature 25℃, no external interference). When it fails, it prompts "Recommended initial and medial combination list" in a fixed order.

19. The method as described in claim 16, characterized in that, Cross-border communication scenarios support bidirectional offline translation between Chinese and English and languages ​​with tones; the accuracy rate for translating common phrases is ≥92%, the accuracy rate for translating professional terms is ≥88%, and the translation response time is ≤300ms; single-key dual-phoneme combination rules are reused to languages ​​with tones, and the conflict rate of bilingual mixed input with tones is ≤3%.

20. The method as described in claim 16, characterized in that, In industrial control scenarios, core instructions support offline caching (cache quantity ≥ 1000, retention time ≥ 7 days), and the input time of core instructions is ≤ 1 second; after integrating single-key dual-phoneme combination, the input time of instruction with tone is ≤ 180ms / instruction, the offline cache combination instruction is ≥ 800 groups, and the availability after network disconnection is ≥ 99.8%.

21. A single-key dual-phoneme enhanced human-computer interaction method based on the method of claim 1, characterized in that, include: 1) Single-key double-phoneme combination type: ① Medial vowel and number single tone / character tone combination (i + 1-4 tone / neutral tone, u / ü + 1-4 tone / neutral tone); ② Tone and medial vowel combination (1-4 tone / neutral tone + i, 1-4 tone / neutral tone + u / ü, tone order can be adjusted); 2) Single-key triggering logic and indivisibility: All dual-phoneme combinations are triggered by a single key (0 key, bound to the medial i) / "-" key (bound to the medial u / ü). The triggering method includes two irreplaceable implementation paths: ① Long press of a digital key (long press duration ≥ 300ms, customizable threshold range 100-500ms); ② Shortcut key combination triggering: Preset combination keys are locked through system permission requests to prevent third-party software from occupying them. Triggering occurs directly through the system's underlying interface, without being intercepted by third-party software. Dual-key input is triggered by a short press (short press duration...). The system distinguishes the triggering method by dual judgment of press duration and signal characteristics (≤200ms), with no conflict and a trigger response delay of ≤50ms; and the encoding sequence of single-key dual-phoneme combination forms a unique correspondence with the core encoding system of number sequence vowels, which cannot be replaced by equivalent means; the aforementioned "single-key triggering-encoding mapping-cross-mode compatibility" constitutes an indivisible functional whole, and it is prohibited to circumvent it by splitting "triggering operation and encoding generation" into independent steps. That is, any scheme that first triggers the key and then generates an equivalent dual-phoneme code through an additional software module falls within the protection scope of this method; 3) Cross-mode compatibility: The single-key dual-phoneme combination rule applies to the main tone set, consonant set and all Pinyin input modes, and works in conjunction with the dual-mode tone function.

22. The method as described in claim 21, characterized in that, The tone triggering methods for single-key dual-phoneme combinations are as follows: single-tone numbers are triggered by numbers 1-4 and 5; dual-tone numbers are triggered by numbers 6, 7-0; and character tones are triggered by independent keys or the "punctuation-tone" shared key.

23. The method as described in claim 21, characterized in that, Supports industry-specific extensions: Medical, financial, and industrial fields can import their own dual-phoneme combination rule bases, and the system automatically associates industry-specific tone combinations; professional terminology tone input response latency ≤70ms, annotation accuracy ≥99%.

24. A multi-input mode human-computer interaction method based on the method described in claim 1 or 21, characterized in that, include: 1) Full coverage of input modes: It includes six input modes: abbreviated pinyin, abbreviated pinyin extension, superimposed pinyin, moving double pinyin, single-key full pinyin, and double-key full pinyin. All modes are constructed based on the number sequence vowels described in claim 1. 2) Core functions of each mode: ① Simplified Pinyin: Supports encoding of Chinese and English initials / core syllables, bilingual mixed encoding, and index response latency ≤200ms; ②Simplified Pinyin Extension: Superimposes the logic of "medial vowel + single-key number sequence vowel". The medial vowel can be triggered by either a long press on the soft keyboard to open a pop-up window / directional swiping or a long press on the hard keyboard to activate and a short press to output. ③ Overlay spelling: Triggered by the initial and medial vowels, compatible with abbreviated spelling, abbreviated spelling extension and full spelling modes, and supports mode conversion (accuracy ≥99%); ④ Mobile Double Pinyin: Lowercase / uppercase initials and English characters are triggered by double keys; medial vowels with zero medial vowels are triggered directly by number-order vowels; reusing the number-order dual-tone system in the number-order tone extension method of claim 7 and the single-key dual-phoneme combination rule of claim 21, after inputting an initial + number-order vowel, the system automatically triggers single-key dual-phoneme combination through the 0 key (bound to medial i) and the "-" key (bound to medial u / ü), directly prompting the simplified code word or syllable symbol of the initial-medial-vowel-tone syllable containing the medial vowel, with a combination response delay ≤50ms (test environment: room temperature 25℃±5℃, device memory ≥2GB); simplified code words or syllable symbols of the initial-medial-vowel-tone syllable without the medial vowel can be input through the consonant set character tone input of claim 10 (including independent tone keys or "punctuation-tone" shared key keys), or reuse the rights claim For inputting the 7-digit number-tone system, users can choose between two methods, sharing the tone-tone combination vocabulary mapping library. The 0 key corresponds to i + number-order vowel + tone, and the "-" key corresponds to u / ü + number-order vowel + tone. Furthermore, the mobile double-pinyin mode has built-in splitting avoidance identification logic: for splitting behaviors such as "splitting the initial consonant input and the medial vowel trigger as two independent software calls" and "splitting the encoding generation and candidate prompts as different modules," it automatically identifies and rejects equivalent inputs by detecting the temporal correlation of the input signal (the interval between consecutive operations within the same input session is ≤100ms) and the module call chain. At the same time, a clear splitting avoidance verification standard is defined: if the input steps, encoding generation logic, and candidate output results of a certain scheme are ≥90% equivalent to the core features of this mode, and are achieved through splitting technology features, then it is determined to be an infringement. ⑤ Single-key full pinyin input: Adapts to existing technology comparison scenarios, allowing users to input pinyin characters one by one; ⑥ Double-key full spelling: Corresponds to the single-key full spelling logic of the letter keyboard, suitable for children's basic language teaching scenarios; 3) Mode Coordination Switching: Supports automatic switching based on input sequence or manual switching by the user, with a switching delay of ≤50ms.

25. The method as described in claim 24, characterized in that, The triggering rules for the soft keyboard of the simplified pinyin extension mode are as follows: press and hold the 0 key / "-" key for ≥300ms (≥500ms in children's mode) to pop up a large candidate pop-up window, or "middle key + directional slide", with a recognition accuracy of ≥96%, and simultaneously display the "middle + tone" single key dual phoneme combination.

26. The method as described in claim 24, characterized in that, The mobile double-pinyin hard keyboard blind typing rules include: allowing users to adjust the combination order; skilled users achieve a blind typing accuracy of ≥95%; supporting the binding of single-key dual-phoneme combination shortcut keys; and a response latency of ≤15ms.

27. A multilingual human-computer interaction system based on the method of any one of claims 1-26, characterized in that, include: 1) Input module: Receives user key operations, voice or gesture input, and outputs input signals with feature identifiers (key input signals include press duration / force features, and voice input signals include voiceprint / syllable features); 2) Encoding Module: Pre-stores unmodifiable numerical vowel mapping rules and mother-child interface configurations, converts the input signal into corresponding initial-medial-final codes or multilingual character codes, and has a built-in encoding conflict detection unit. The rejection response delay for invalid encoding combinations (such as "zh+ü" "b+iong") is ≤150ms (test environment: room temperature 25℃±5℃, humidity 40%-60%, no external electromagnetic interference); the numerical vowel mapping rules and mother-child interface configurations correspond to the methods described in any one of claims 1-26, and are prevented from being tampered with by a hardware chip-level encryption unit; 3) Candidate processing module: Based on the encoding, the encrypted offline dictionary is called to generate a candidate character / vocabulary list, which supports dual sorting by input frequency and semantic relevance, and the candidate list update delay is ≤30ms; 4) Display module: Displays the vowel interface, candidate list and extended interface. The interface rendering follows the principle of "horizontal alignment between the trigger area and the candidate area", with an alignment error of ≤1 pixel. It supports adaptive screen size of multiple devices (adaptation range: 1.5 inches-21 inches). 5) Control Module: Employing a multi-threaded collaborative architecture, the module controls the synchronous operation of the above modules, enabling input mode switching, multimodal interaction, and cross-device adaptation with a switching latency of ≤50ms. It includes a built-in anti-circumvention monitoring unit, focusing on scenarios such as technical feature splitting and device-specific circumvention. It identifies and intercepts such actions through the following rules: ① Monitoring module call chains and rejecting input behaviors that split "encoding conversion-conflict detection-candidate generation" into independent modules without collaborative verification; ② Detecting the temporal continuity of input signals and intercepting behaviors where splitting core operations (such as splitting key triggers and encoding mappings) is a non-continuous operation; ③ Verifying the integrity of core technical features and directly rejecting and recording circumvention behaviors that only call some modules (such as only calling the input module and display module, skipping the encryption verification of the encoding module) to achieve equivalent input; ④ For small-screen devices such as smartwatches, intercepting circumvention behaviors such as "using only '11' and '12' extended keys but deviating from the core mapping table" and "modifying the key-vowel correspondence of small-screen devices." 6) Hardware encryption unit: Integrated with the encoding module, it performs hardware-level encryption storage of core mapping rules and offline dictionaries, allowing reading only through authorized interfaces and prohibiting modification; the encryption algorithm adopts the Chinese national standard SM4 and the international standard AES-256, supports automatic switching of encryption standards according to the compliance requirements of the target country, and generates encryption compliance reports that conform to local standards, adapting to the encryption compliance requirements of different countries.

28. The system as claimed in claim 27, characterized in that, It also includes an intelligent switching module: automatically switching input modes based on user input sequences or multimodal interaction features. When inputting "single / double-key initials", it enters the stacked and compatible simplified pinyin mode; when inputting "initial + medial", it switches to the simplified pinyin extended mode; when inputting "initial, medial, and double-key", it enters the stacked pinyin dedicated mode; when triggering tone input, it switches to the corresponding tone mode; when triggering multimodal input, it switches to the corresponding interaction mode; and when a single-key dual-phoneme encoding sequence is detected, it automatically switches to the corresponding mode; the switching delay is ≤50ms.

29. The system as claimed in claim 27, characterized in that, It also includes an AI collaboration module: interacting with generative AI large models, realizing semantic-level input completion and multi-turn dialogue encoding collaboration functions based on number-order vowel encoding logic, with an AI collaboration accuracy of ≥92%; recommending high-frequency single-key dual-phoneme combinations based on input context, with a word completion accuracy of ≥92% with tone.

30. The system as claimed in claim 27, characterized in that, Supports Android 8.0+, iOS 12.0+, Windows 10+, Linux (Ubuntu 18.04+), and industrial real-time operating systems (VxWorks 7.0+, QNX 7.0+); cross-platform compatibility ≥99%, installation package size: basic version ≤20MB, full version ≤50MB; and supports compliant adaptation of configuration parameters for encoding modules and encryption units to meet the patent examination requirements of different countries, satisfying the data security and encryption standards of various countries without changing the core claims; furthermore, the system has a patent infringement early warning function, which can monitor and record counterfeit use of core encoding rules and parent-child interface logic, and generate an infringement evidence chain containing timestamps and device identifiers; a new technical feature splitting and circumvention verification scheme has been added, specifically including: 1) Test scenario: simulate splitting the "three-in-one mapping table" of claim 1 (only implementing vowel-key mapping, skipping encoding sequence binding), splitting the "..." of claim 21 1) Single-key dual-phoneme triggering logic (first triggering the key and then generating dual-phoneme encoding through a third-party module), splitting the "module collaboration logic" of claim 27 (only calling some modules to achieve equivalent input); 2) Test method: Under a standard test environment (room temperature 25℃±5℃, humidity 40%-60%, device memory ≥2GB), the above splitting avoidance input behavior is generated through an automated test tool, and the system response results are recorded; 3) Judgment criteria: If the system can 100% identify and intercept the above splitting avoidance behavior and generate a warning log containing "split behavior type, trigger time, device identifier", it is judged as passing the verification; if there is uninterrupted splitting avoidance behavior, it is judged that the core technical features are not effectively protected, and anti-avoidance logic optimization needs to be started.

Images