A multi-lingual intelligent interaction processing system

Abstract

This invention discloses a multilingual intelligent interactive processing system, relating to the field of data processing, comprising: a collection module for collecting text speech data sent by each interactive subject in a group chat scenario, and simultaneously acquiring the unique identifier of the corresponding speaker, the speech time sequence marker, and the subject's interface language configuration information; and a statistics module for sorting the collected text speech data according to the speech time sequence, and simultaneously calculating the interface language type and total number of all interactive subjects in the group chat. This invention performs real-time intelligent adaptation processing for multilingual interaction in group chat scenarios, accurately matches the interface language configuration of different users, automatically completes the multilingual synchronous conversion of speech content, breaks down language communication barriers in group chats, and allows users of different languages ​​to receive native-language interactive information, thereby improving the communication efficiency of cross-language group chats.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a multilingual intelligent interactive processing system. Background Technology

[0002] Multilingual interaction refers to a system's ability to support input and output in multiple languages, enabling natural cross-language communication and information exchange. It integrates technologies such as speech recognition, machine translation, and natural language understanding, accurately understanding user intent in different language scenarios and responding in the target language. It is widely used in intelligent customer service, cross-border communication, and online service scenarios.

[0003] Patent application No. 202510176854.3 discloses a text-based dialogue method, apparatus, terminal device, and storage medium applicable to multiple languages. This application aims to address the problem that "traditional text-based dialogue technologies, when processing user-input text, directly perform contextual analysis on the source text and attempt to understand the user's intent and contextual information. Based on the results of the contextual analysis, a response text is generated. However, traditional dialogue technologies cannot identify the language of the input text, which may lead to responses that do not match the user's expected language. Furthermore, because the language of the current input text is not considered, responses cannot be made based on the actual grammatical logic of that language, resulting in grammatical errors, awkwardness, or incoherence in the response text. This leads to low accuracy and naturalness in responses during the dialogue process, and makes multilingual text-based dialogue impossible."

[0004] However, in existing multilingual text dialogue scenarios, most interactions are one-to-one. When faced with text interaction scenarios in group chats where there are two or more languages, there are no existing technologies adapted to this type of scenario to process and manage the interactive information.

[0005] To address this, we propose a multilingual intelligent interactive processing system. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a multilingual intelligent interactive processing system that can effectively solve the problems of the existing technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions; This invention discloses a multilingual intelligent interactive processing system, comprising: a collection module for collecting text speech data sent by each interactive subject in a group chat scenario, and simultaneously acquiring the unique identifier of the corresponding speaker, the speech sequence marker, and the subject's interface language configuration information; a statistics module for sorting several collected text speech data according to the speech sequence, and simultaneously calculating the interface language types and total number of all interactive subjects in the group chat; a construction module for constructing a dedicated language adaptation database for a single group chat scenario based on the statistically obtained interface display language types and total number; a conversion module for performing synchronous conversion processing on a single text speech data in the dedicated language adaptation database according to the speech sequence sorting results, based on the constructed dedicated language adaptation database; an encapsulation module for matching the unique identifier of the receiving subject in the corresponding interface display language to each target language text data after language conversion, encapsulating the speaker identifier and the speech sequence marker, and generating a single subject adaptation data packet; and a distribution module for distributing the encapsulated single subject adaptation data packet to the terminal device of the corresponding interactive subject, and displaying it in the group chat window on the terminal device. The data acquisition module is interconnected with the statistics module via a wireless network. The data acquisition module and the statistics module are interconnected with the construction module via a wireless network. The construction module is interconnected with the conversion module, the packaging module, and the distribution module via a wireless network. The conversion module, the packaging module, and the distribution module are interconnected with each other via a wireless network.

[0008] Furthermore, during the data acquisition module's operation phase, it synchronously identifies reference association markers and directional interaction markers in the text speech data, identifies the unique identifier of the speaker subject corresponding to the referenced speech and the speech timing marker, generates an associated speech index corresponding to a single text speech data, and binds and stores the associated speech index with the corresponding text speech data, the unique identifier of the speaker subject, the speech timing marker, and the main interface language configuration information.

[0009] Furthermore, during the statistical module's operation phase, clock synchronization calibration is performed on the speech timing markers of the collected text speech data to eliminate clock deviations between different terminal devices; the calibrated speech timing markers conform to: In the formula, The calibrated speech timing marker for the i-th text speech data; The original speech time sequence markers collected from the i-th text speech data; The clock deviation between the terminal device sending the i-th text message and the system reference clock; The system presets a global timing compensation threshold; after completing clock synchronization calibration, the statistics module counts the interface language types corresponding to all interactive subjects in the group chat scenario based on the calibrated speech timing markers, and counts the number of interactive subjects in a valid receiving state corresponding to each interface language type.

[0010] Furthermore, based on the statistically obtained interface display language types and total number, the construction module first extracts bidirectional language conversion rules corresponding one-to-one with all interface display languages ​​from the system's pre-set set of basic language conversion rules, generating a set of basic conversion rules adapted to the current group chat scenario. Then, it extracts historical text speech data within the current group chat scenario for a preset statistical period, identifies exclusive words and fixed expressions that meet the preset frequency threshold and are not included in the set of basic language conversion rules for fixed translations. For each identified exclusive word and fixed expression, it matches a unique and fixed translation corresponding to all interface display languages ​​obtained from statistics, generating a scene terminology mapping table specific to the group chat. The basic conversion rule set and the scene terminology mapping table are then structurally integrated to construct a specific language adaptation database for a single group chat scenario. At the same time, within the specific language adaptation database, the conversion execution priority of the scene terminology mapping table is set to be higher than that of the basic conversion rule set.

[0011] Furthermore, during the conversion module's execution phase, based on the associated speech index corresponding to the text speech data, the original content of the associated speech and the corresponding language conversion result are extracted. The original content of the current text speech data is then combined to perform context-dependent language conversion processing. Simultaneously, cross-language semantic consistency checks are performed on the text data in different target languages ​​obtained from the conversion of the same original content. ;in, The semantic deviation between the source language text and the target language conversion result; This is a function for calculating semantic similarity within a vector space. This is the semantic encoding vector corresponding to the source language text data; This is the semantic encoding vector corresponding to the text data converted to the target language; when the semantic deviation exceeds the preset deviation threshold, the conversion optimization and correction of the corresponding target language text data is triggered.

[0012] Furthermore, during the conversion module's operation phase, based on the calibrated speech timing markers, a unified timing lock identifier is bound to all target language conversion tasks corresponding to the same text speech data. Only when all target language conversion tasks corresponding to the speech timing marker are completed will the conversion-completed target language text data be output to the encapsulation module.

[0013] Furthermore, when the encapsulation module runs, it matches the target language text data that has been converted to the target language with the unique identifiers of all receiving entities whose interface language is configured to the target language, establishes a mapping relationship between a single target language text data and the corresponding set of unique identifiers of receiving entities, and then performs end-to-end encryption processing on the adaptation data packet corresponding to each unique identifier of receiving entity, embedding a check code in the data packet.

[0014] Furthermore, the distribution module allocates corresponding distribution channels and retransmission mechanisms for data packets adapted to a single subject based on the online status and network status of the receiving subject's terminal device. When a data packet distribution failure is detected, a retransmission operation is performed based on a preset retransmission strategy, and the terminal device is simultaneously controlled to display the corresponding converted target language text data in the group chat window according to the order of the calibrated speaking time sequence markers, and the speaking subject identification information is displayed synchronously.

[0015] Furthermore, each module in the system performs monitoring of changes in interactive subjects within the group chat scenario during its operation. When an increase, decrease, or replacement of interactive subjects is detected within the group chat scenario, the system is triggered to refresh and update the language adaptation database specific to the group chat scenario.

[0016] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects: This invention provides real-time intelligent adaptation processing for multilingual interaction in group chat scenarios, accurately matching the interface language configuration of different users, automatically completing the multilingual synchronous conversion of the speech content, breaking down language communication barriers in group chats, and allowing users of different languages ​​to receive native-language interactive information, thereby improving the communication efficiency of cross-language group chats. In this invention, the system constructs a language adaptation rule and scenario terminology library specifically for group chats, prioritizing the adaptation of exclusive vocabulary and fixed expressions to ensure that the converted content fits the actual use scenario of group chats, resulting in higher conversion accuracy. After clock synchronization calibration, the speaking sequence is unified to avoid information confusion caused by terminal clock deviation. Combined with context-related conversion and semantic consistency verification, the semantics of multilingual conversion results are ensured to be accurate and consistent. The system performs targeted distribution and end-to-end encryption of the adaptation data packets, taking into account both the security and delivery rate of information transmission. When group chat members change, the system automatically updates to complete the adaptation, without the need for manual intervention throughout the entire process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 This is a schematic diagram of the structure of a multilingual intelligent interactive processing system; Figure 2 A logical example diagram of generating a single-subject adaptation data packet for the encapsulation module in the invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] The present invention will be further described below with reference to embodiments.

[0021] Example: This embodiment provides a multilingual intelligent interactive processing system, such as... Figure 1 As shown, it includes: The data collection module is used to collect text speech data sent by each interactive subject in the group chat scenario, and simultaneously obtain the unique identifier of the corresponding speaker, the speech time sequence mark and the language configuration information of the subject interface; During the data acquisition module's operation phase, it synchronously identifies reference association markers and directional interaction markers in the text speech data, marks the unique identifier of the speaker subject and the speech time sequence marker of the corresponding referenced speech, generates the associated speech index corresponding to a single text speech data, and binds and stores the associated speech index with the corresponding text speech data, the unique identifier of the speaker subject, the speech time sequence marker, and the main interface language configuration information; The statistics module is used to sort the collected text messages according to the speaking time, and simultaneously count the interface language type and total number of all interactive subjects in the group chat. During the operation of the statistics module, clock synchronization calibration is performed on the speech time sequence markers of the collected text speech data to eliminate clock deviations between different terminal devices; The calibrated speech timing markers conform to: ; In the formula: The calibrated speech timing marker for the i-th text speech data; The original speech time sequence markers collected from the i-th text speech data; The clock deviation between the terminal device sending the i-th text message and the system reference clock; The global timing compensation threshold preset for the system; The above formula completes the synchronous calibration of the group chat speaking time sequence by combining the original speaking time sequence, device clock deviation and system global timing compensation value, eliminates the clock difference between different terminals, makes the group chat speaking time sequence a unified benchmark, accurately matches the actual speaking order, and thus ensures the timing accuracy of multilingual conversion and message distribution. After completing clock synchronization calibration, the statistics module counts the interface language types corresponding to all interactive subjects in the group chat scenario based on the calibrated speech timing markers, and counts the number of interactive subjects in a valid receiving state corresponding to each interface language type. The module is used to build a dedicated language adaptation database for single and group chat scenarios based on the statistically obtained interface display language types and total numbers. Based on the statistically obtained interface display language types and total number, the construction module first extracts bidirectional language conversion rules corresponding to all interface display languages ​​from the system's pre-set set of basic conversion rules for all languages, generating a set of basic conversion rules adapted to the current group chat scenario. Then, it extracts historical text speech data within the current group chat scenario for a preset statistical period, identifies exclusive words and fixed expressions that meet the preset frequency threshold and are not included in the set of basic conversion rules for all languages ​​with fixed translations. For each identified exclusive word and fixed expression, it matches the unique and fixed translations corresponding to all interface display languages ​​obtained from the statistics, generating a scene terminology mapping table specific to the group chat. The basic conversion rule set and the scene terminology mapping table are then structurally integrated to construct a specific language adaptation database for a single group chat scenario. At the same time, within the specific language adaptation database, the conversion execution priority of the scene terminology mapping table is set to be higher than that of the basic conversion rule set. The conversion module is used to perform synchronous conversion processing on a single text speech data in all languages ​​except the language of the speech data, based on the constructed dedicated language adaptation database and the speech time order. During the conversion module's operation, based on the associated speech index corresponding to the text speech data, the original content of the associated speech and the corresponding language conversion result are extracted. Then, the original content of the current text speech data is combined to perform context-dependent language conversion processing. Simultaneously, cross-language semantic consistency checks are performed on the text data in different target languages ​​obtained from the same original content. ; In the formula: The semantic deviation between the source language text and the target language conversion result; This is a function for calculating semantic similarity within a vector space. This is the semantic encoding vector corresponding to the source language text data; This is the semantic encoding vector corresponding to the text data after conversion to the target language; Based on the semantic similarity calculation results, this formula subtracts the similarity value from 1 to obtain the semantic deviation between the source language and the target language. It can intuitively quantify the semantic differences after translation, quickly identify deviations and trigger optimization and correction, ensure the semantic consistency of different language texts after multilingual conversion, and improve the accuracy of group chat interaction understanding. Specifically, when the semantic deviation exceeds a preset deviation threshold, the conversion optimization and correction of the corresponding target language text data is triggered. Choose any existing semantic similarity calculation function that is suitable for the current scenario; During the operation phase of the conversion module, based on the calibrated speech timing mark, a unified timing lock identifier is bound to all target language conversion tasks corresponding to the same text speech data. Only when all target language conversion tasks corresponding to the speech timing mark are completed will the conversion of each target language text data completed by the conversion be output to the encapsulation module. The encapsulation module is used to match the unique identifier of the receiving subject in the corresponding interface display language to the target language text data after language conversion, encapsulate the speaker identifier and the speaking time sequence mark, and generate a single subject adaptation data packet. When the encapsulation module runs, it will match the target language text data that has been converted to the target language with the unique identifiers of all receiving entities whose interface language is configured to the target language, establish a mapping relationship between a single target language text data and the set of corresponding unique identifiers of receiving entities, and then perform end-to-end encryption processing on the adaptation data packet corresponding to each unique identifier of receiving entity, embedding a check code in the data packet. The encryption algorithm used in the encryption process is preset on the system side, and the verification code is generated based on the unique identifier of the speaker, the speaking time sequence mark and the target language text data. The distribution module is used to distribute the encapsulated single-subject adaptation data package to the terminal device of the corresponding interactive subject and display it in the group chat window on the terminal device. Based on the online status and network status of the receiving entity's terminal device, the distribution module allocates the corresponding distribution channel and retransmission mechanism for the data packets adapted to a single entity. When a data packet distribution failure is detected, a retransmission operation is performed based on the preset retransmission strategy. Simultaneously, the terminal device is controlled to display the corresponding converted target language text data in the group chat window according to the order of the calibrated speaking time sequence markers, and the speaking entity identification information is displayed in conjunction with the data. Among them, terminal devices are the interactive devices used by each interactive subject to perform interactive operations in the group chat scenario, such as mobile phones, laptops, and computers; During the operation of each module in the system, the system performs monitoring of changes in interactive subjects within the group chat scenario. When the system detects an increase, decrease, or replacement of interactive subjects within the group chat scenario, it triggers a system refresh to update the database for language adaptation specific to the group chat scenario. The data acquisition module is interconnected with the statistics module via a wireless network. The data acquisition module and the statistics module are interconnected with the construction module via a wireless network. The construction module is interconnected with the conversion module, the packaging module, and the distribution module via a wireless network. The conversion module, the packaging module, and the distribution module are interconnected with each other via a wireless network.

[0022] In this embodiment, the acquisition module collects text messages sent by each interactive subject in the group chat scenario, and simultaneously obtains the unique identifier of the corresponding speaker, the message timing mark, and the subject's interface language configuration information. The statistics module simultaneously sorts the collected text messages according to the message timing and simultaneously counts the interface language types and total number of all interactive subjects in the group chat. The construction module constructs a dedicated language adaptation database for the corresponding single group chat scenario based on the statistically obtained interface display language types and total number. The conversion module runs based on the constructed dedicated language adaptation database and performs synchronous conversion processing on all languages ​​in the dedicated language adaptation database except for the message language, according to the message timing sorting results. Then, the encapsulation module matches the unique identifier of the receiving subject in the corresponding interface display language for each target language text data that has completed language conversion, encapsulates the speaker identifier and the message timing mark, and generates a single subject adaptation data packet. Finally, the distribution module distributes the encapsulated single subject adaptation data packet to the terminal device of the corresponding interactive subject and displays it in the group chat window on the terminal device.

[0023] In this embodiment, when the system is applied to group chat interaction scenarios, it can automatically adapt to the interface language of all members in the group, complete the multilingual synchronous conversion of the speech content in real time, accurately match the reading needs of each member, and ensure the accuracy of the speech display order through time sequence calibration. It also automatically updates the adaptation rules when members change, so that users of different languages ​​can communicate smoothly without manually switching, thereby improving the efficiency of group chat communication and user experience.

[0024] See Figure 2 As shown in the diagram, there are five users, A, B, C, D, and E, in this example group chat. The five users speak one after another at the same time. Based on the arrow indication, the data is decomposed into five groups of single-subject adaptation data packets. Each group of single-subject adaptation data packets is directed to the other four users and contains four single-subject adaptation data packets. Finally, the data packet groups are synchronously distributed to each user's receiving end (each user receives the single-subject adaptation data packets from each user in their own receiving end).

[0025] Referring to the system in the above embodiments, the following is an application example of the system: A multinational company has set up a business collaboration group chat that includes Chinese, English, and Japanese users. After the system is enabled, real-time interaction within the group is achieved without language barriers.

[0026] The system's data acquisition module captures the text of messages sent by users in the group in real time, and simultaneously obtains each user's unique identifier, message time, and interface language configuration. It also identifies quotations and targeted interaction markers in the messages, generates associated message indexes, and binds and stores them.

[0027] The statistics module completes clock synchronization calibration for all speaking times, obtaining a unified and accurate calibrated speaking time sequence. It also identifies the languages ​​of the effective users in the group as Chinese, English, and Japanese, with corresponding user numbers of 5, 3, and 2 respectively.

[0028] The module extracts bidirectional conversion rules for three types of languages ​​from the system's basic rules, then extracts high-frequency business-specific vocabulary within the group to generate a scenario terminology mapping table. After integration, it builds a language-specific adaptation database for the group chat, with the conversion priority of specific vocabulary being higher than that of the basic rules.

[0029] The conversion module converts each message into two other languages ​​in sync according to the calibrated time sequence, performs context-related conversion with related messages, and verifies cross-language semantic consistency. The semantic deviation is within a preset reasonable range and does not require correction. Only after all languages ​​of a single message have been converted, does it proceed to the next process.

[0030] The encapsulation module matches the converted text in each language with the unique identifier of the receiving user in the corresponding language, binds the speaker identifier and timing mark to generate a single-user adapted data packet, performs end-to-end encryption on the data packet and embeds a checksum.

[0031] The distribution module allocates distribution channels based on the user terminal's online and network status, accurately targets and sends data packets, and automatically resends them if distribution fails; the terminal group chat window displays the content of the messages and the information of the speakers in the corresponding languages ​​according to the calibrated time sequence.

[0032] If a new Korean user is added or a Japanese user is removed from the group, the system will detect the member changes in real time, automatically refresh and update the dedicated language adaptation database to adapt to the new language interaction needs.

[0033] Overall, in the above embodiments, the system performs real-time intelligent adaptation processing for multilingual interaction in group chat scenarios, accurately matches the interface language configuration of different users, automatically completes the multilingual synchronous conversion of the speech content, breaks down language communication barriers in group chats, and allows users of different languages ​​to receive native language interactive information, thereby improving the communication efficiency of cross-language group chats. During system operation, by constructing a language adaptation rule and scenario terminology library specific to group chats, priority is given to adapting exclusive vocabulary and fixed expressions to ensure that the converted content fits the actual use scenario of group chats, resulting in higher conversion accuracy. After clock synchronization calibration, the speaking sequence is unified to avoid information confusion caused by terminal clock deviation. Combined with context-related conversion and semantic consistency verification, the semantics of multilingual conversion results are ensured to be accurate and consistent. Adaptation data packets are distributed in a targeted manner and encrypted end-to-end to balance the security and delivery rate of information transmission. When group chat members change, the system automatically updates to complete the adaptation, without the need for manual intervention throughout the process.

[0034] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multilingual intelligent interactive processing system, characterized in that, include: The data collection module is used to collect text speech data sent by each interactive subject in the group chat scenario, and simultaneously obtain the unique identifier of the corresponding speaker, the speech time sequence mark and the language configuration information of the subject interface; The statistics module is used to sort several pieces of text speech data according to the speaking time sequence, and simultaneously count the interface language type and total number of all interactive subjects in the group chat. The module is used to build a dedicated language adaptation database for single and group chat scenarios based on the statistically obtained interface display language types and total numbers. The conversion module is used to perform synchronous conversion processing on a single text speech data in all languages ​​except the language of the speech data, based on the constructed dedicated language adaptation database and the speech time order. The encapsulation module is used to match the unique identifier of the receiving subject in the corresponding interface display language to the target language text data after language conversion, encapsulate the speaker identifier and the speaking time sequence mark, and generate a single subject adaptation data packet. The distribution module is used to distribute the encapsulated single-subject adaptation data package to the terminal device of the corresponding interactive subject and display it in the group chat window on the terminal device.

2. The multilingual intelligent interactive processing system according to claim 1, characterized in that, During the operation phase of the acquisition module, the reference association markers and directional interaction markers in the text speech data are identified simultaneously. The unique identifier of the speaker and the speech time sequence marker of the corresponding referenced speech are marked. The associated speech index corresponding to a single text speech data is generated, and the associated speech index is bound and stored with the corresponding text speech data, the unique identifier of the speaker, the speech time sequence marker, and the main interface language configuration information.

3. The multilingual intelligent interactive processing system according to claim 1, characterized in that, During the operation phase of the statistical module, clock synchronization calibration is performed on the speech timing markers of the collected text speech data to eliminate clock deviations between different terminal devices; The calibrated speech timing markers conform to: ； In the formula: The calibrated speech timing marker for the i-th text speech data; The original speech time sequence markers collected from the i-th text speech data; The clock deviation between the terminal device sending the i-th text message and the system reference clock; The global timing compensation threshold preset for the system; After completing clock synchronization calibration, the statistics module counts the interface language types corresponding to all interactive subjects in the group chat scenario based on the calibrated speech timing markers, and counts the number of interactive subjects in a valid receiving state corresponding to each interface language type.

4. The multilingual intelligent interactive processing system according to claim 1, characterized in that, The construction module, based on the statistically obtained interface display language types and total number, first extracts bidirectional language conversion rules corresponding one-to-one with all interface display languages ​​from the system's pre-set set of basic language conversion rules, generating a set of basic conversion rules adapted to the current group chat scenario. Then, it extracts historical text speech data within a preset statistical period in the current group chat scenario, identifies exclusive words and fixed expressions whose frequency meets a preset frequency threshold and are not included in the set of basic language conversion rules for fixed translations. For each identified exclusive word and fixed expression, it matches a unique and fixed translation corresponding to all interface display languages ​​obtained from statistics, generating a scene terminology mapping table specific to that group chat. The basic conversion rule set and the scene terminology mapping table are then structurally integrated to construct a dedicated language adaptation database for a single group chat scenario. Simultaneously, within the dedicated language adaptation database, the conversion execution priority of the scene terminology mapping table is set higher than that of the basic conversion rule set.

5. The multilingual intelligent interactive processing system according to claim 2, characterized in that, During the operation of the conversion module, based on the associated speech index corresponding to the text speech data, the original content of the associated speech and the corresponding language conversion result are extracted. Then, the original content of the current text speech data is combined to perform context-dependent language conversion processing. Simultaneously, cross-language semantic consistency checks are performed on the text data in different target languages ​​obtained from the conversion of the same original content. ； In the formula: The semantic deviation between the source language text and the target language conversion result; This is a function for calculating semantic similarity within a vector space. This is the semantic encoding vector corresponding to the source language text data; This is the semantic encoding vector corresponding to the text data after conversion to the target language; Specifically, when the semantic deviation exceeds the preset deviation threshold, the conversion optimization and correction of the corresponding target language text data is triggered.

6. The multilingual intelligent interactive processing system according to claim 3, characterized in that, During the operation phase of the conversion module, based on the calibrated speech timing marker, a unified timing lock identifier is bound to all target language conversion tasks corresponding to the same text speech data. Only when all target language conversion tasks corresponding to the speech timing marker are completed will the module output the target language text data with completed language conversion to the encapsulation module.

7. The multilingual intelligent interactive processing system according to claim 1, characterized in that, When the encapsulation module is running, it matches the target language text data that has been converted to the target language with the unique identifiers of all receiving entities whose interface language is configured to the target language, establishes a mapping relationship between a single target language text data and the corresponding set of unique identifiers of receiving entities, and then performs end-to-end encryption processing on the adaptation data packet corresponding to each unique identifier of receiving entity, embedding a check code in the data packet.

8. The multilingual intelligent interactive processing system according to claim 1, characterized in that, The distribution module allocates corresponding distribution channels and retransmission mechanisms for data packets adapted to a single subject based on the online status and network status of the receiving subject's terminal device. When a data packet distribution failure is detected, a retransmission operation is performed based on a preset retransmission strategy. Simultaneously, the terminal device is controlled to display the corresponding converted target language text data in the group chat window according to the order of the calibrated speaking time sequence markers, and the speaking subject identification information is displayed synchronously.

9. A multilingual intelligent interactive processing system according to claim 1, characterized in that, Each module in the system performs monitoring of changes in interactive subjects within the group chat scenario during its operation phase. When an increase, decrease, or replacement of interactive subjects is detected within the group chat scenario, the system is triggered to refresh and update the language adaptation database specific to the group chat scenario.

10. A multilingual intelligent interactive processing system according to claim 1, characterized in that, The data acquisition module is interconnected with the statistics module via a wireless network. The data acquisition module and the statistics module are interconnected with the construction module via a wireless network. The construction module is interconnected with the conversion module, the packaging module and the distribution module via a wireless network. The conversion module, the packaging module and the distribution module are interconnected via a wireless network.

Images