System for automatically generating dialect homophone vocabulary

Abstract

The invention discloses a dialect homophone vocabulary automatic generation system. The system comprises a pronunciation analysis module which is responsible for analyzing pronunciation characteristics of Shaanxi dialects; the vocabulary library module is responsible for storing vocabularies in Shaanxi dialects, and the vocabularies comprise ancient vocabularies and unique vocabularies; the homophone matching module is responsible for searching characters with the same pronunciation in the vocabulary library according to the input pronunciation or text; the grammar analysis module is responsible for analyzing a grammar structure of the Shaanxi dialect and ensuring that the generated homophone vocabulary conforms to a grammar rule of the Shaanxi dialect; the user interface module is responsible for providing user input and output interfaces including voice input, text input and homophone vocabulary display; in the invention, through cooperative work of the pronunciation analysis module, the homophone matching module and the grammar analysis module, the system can accurately generate the homophone vocabulary conforming to the Shaanxi dialect grammar rule, and the accuracy and reliability of the result are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of word processing, and in particular to a system for automatically generating dialect homophone words. Background Art

[0002] As one of the important birthplaces of Chinese culture, Shaanxi has rich language resources and a wide variety of dialects. The dialects of northern Shaanxi, Guanzhong, and southern Shaanxi are significantly different, and various dialects are constantly changing and developing in mutual influence and penetration. This diversity of dialects provides rich materials for language research, but it also increases the difficulty of dialect investigation and compilation. The automatic generation system of dialect homophones can efficiently generate dialect homophones, providing a convenient tool for dialect researchers. This helps to deeply explore the phonetic system, vocabulary characteristics, grammatical structure, etc. of dialects, and promote the in-depth development of dialect research.

[0003] Shaanxi dialect has a long history and a vast territory, forming a variety of dialect variants with local characteristics. These dialects are significantly different from Mandarin in pronunciation, intonation, vocabulary, etc., and the internal differences are also large. A special system is needed to organize and generate homophones of Shaanxi dialect. Therefore, a system for automatically generating dialect homophones is proposed. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art that the Shaanxi dialect has significant differences from Mandarin in pronunciation, intonation, vocabulary, etc. and the internal differences are also large, and the traditional dialect homophone vocabulary automatic generation system cannot effectively identify and generate the shortcomings, and a dialect homophone vocabulary automatic generation system is proposed.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A system for automatically generating dialect homophone vocabulary, comprising:

[0007] Pronunciation analysis module: responsible for analyzing the pronunciation characteristics of Shaanxi dialect, including voiceless sounds, finals, swallowed sounds, long sounds, flat tongue sounds and curled tongue sounds;

[0008] Vocabulary database module: responsible for storing Shaanxi dialect vocabulary, including ancient vocabulary and unique vocabulary. By establishing a vocabulary database, a large number of Shaanxi dialect vocabulary are collected, classified and annotated. At the same time, users are allowed to add and edit vocabulary;

[0009] Homophone matching module: responsible for searching for words with the same pronunciation in the vocabulary based on the input pronunciation or text;

[0010] Syntax analysis module: responsible for analyzing the grammatical structure of Shaanxi dialect and ensuring that the generated homophones conform to the grammatical rules of Shaanxi dialect;

[0011] User interface module: responsible for providing user input and output interfaces, including voice input, text input and display of homophones;

[0012] The pronunciation analysis module needs to receive the voice or text input by the user and perform pronunciation analysis on it. The pronunciation analysis module outputs the pronunciation features obtained by the analysis to the homophone matching module, and updates the new pronunciation data or rules to the vocabulary library module. The homophone matching module accesses the vocabulary database in the vocabulary library module to obtain vocabulary information in the Shaanxi dialect. The homophone matching module searches for words with the same pronunciation in the vocabulary database according to the pronunciation features provided by the pronunciation analysis module, and outputs the matching results to the grammar analysis module. The grammar analysis module receives the matching results provided by the homophone matching module, performs grammar analysis and verification on the matching results, and outputs the verified homophone vocabulary to the user interface module.

[0013] The above technical solution further includes:

[0014] Furthermore, the pronunciation analysis module includes a speech recognition unit, a pronunciation feature extraction unit, a pronunciation comparison analysis unit and a result output unit, wherein the speech recognition unit is responsible for converting the input Shaanxi dialect speech into text, the speech recognition unit passes the converted text to the pronunciation feature extraction unit, the pronunciation feature extraction unit is responsible for extracting pronunciation features from the converted text or original speech, the pronunciation feature extraction unit passes the extracted pronunciation feature data to the pronunciation comparison analysis unit, the pronunciation comparison analysis unit is responsible for comparing and analyzing the extracted pronunciation features with standard pronunciation features, the pronunciation comparison analysis unit receives the standard pronunciation feature data from the vocabulary library module, the pronunciation comparison analysis unit passes the analysis result to the result output unit, and the result output unit is responsible for outputting the result of the pronunciation analysis to the homophone matching module.

[0015] Furthermore, the vocabulary library module includes a database unit, a user interface unit, a business logic unit, a data verification unit and a log recording unit. The database unit is responsible for storing data of Shaanxi dialect vocabulary, including fields such as vocabulary ID, classification, pronunciation, explanation, usage examples, etc. The database unit provides operation interfaces such as data query, insertion, update and deletion. The user interface unit is responsible for receiving user input, such as adding new vocabulary, editing existing vocabulary, etc., and passing user requests to the database unit for processing. The business logic unit is responsible for processing the business logic of user requests, such as verifying the validity of user input, generating new vocabulary ID, etc. The data verification unit is responsible for verifying the data entered by the user to ensure the legitimacy and accuracy of the data. The system can check whether the vocabulary classification input by the user is valid, whether the pronunciation conforms to the specification, etc. The log recording unit records the user operation log and the system operation status log to facilitate problem troubleshooting and data analysis, and provides a log query function to allow users to view historical operation records. The user interface unit receives the user input and passes it to the business logic unit for processing. After processing the user request, the business logic unit returns the result to the user interface unit. The business logic unit calls the data operation interface provided by the database unit according to the user request. Before processing the user request, the business logic unit calls the data verification unit to verify the user input. When the log recording unit receives event notifications from other units, it records the corresponding log information.

[0016] Furthermore, the homophone matching module uses the pronunciation features provided by the pronunciation analysis module as query conditions and calls the data query interface of the vocabulary library module. The vocabulary library module returns a vocabulary list that matches the input pronunciation according to the query conditions, filters the returned vocabulary list, removes vocabulary that does not meet the conditions (such as mismatched vocabulary types, incorrect pronunciation annotations, etc.), sorts the filtered vocabulary, and generates a final matching result list.

[0017] Furthermore, the grammatical analysis module includes a grammatical rule base unit, a grammatical analysis unit, a grammatical verification unit and a result optimization and feedback unit. The grammatical rule base unit stores the grammatical rules of Shaanxi dialect, including lexical, syntactic and semantic rules. The grammatical rule base unit provides a rule query interface to allow other units to perform rule query as needed. The grammatical analysis unit uses the rules in the grammatical rule base to perform grammatical analysis on homophones. The grammatical analysis unit checks whether the vocabulary collocation and sentence structure in the homophones conform to the grammatical rules of Shaanxi dialect. The grammatical verification unit performs overall grammatical verification on the homophones to ensure that the entire sentence or vocabulary combination is grammatically correct. The grammatical verification unit provides error prompts and suggestions to help users make corrections. The result optimization and feedback unit optimizes the homophones according to the grammatical analysis results. The grammatical analysis unit receives the homophones generated by the homophone matching module, and performs grammatical analysis and verification on the homophones.

[0018] Furthermore, the user interface module includes an input unit and an output unit, the input unit includes voice input and text input, the voice input is the user inputting the voice of Shaanxi dialect through the microphone of a mobile phone, computer and other devices, and the text input is the user manually inputting the text of Shaanxi dialect, which is convenient for homophone search for specific words, and the output unit includes homophone vocabulary display and export function, the homophone vocabulary display will display the matched homophones to the user in the form of a list, including dialect vocabulary, Mandarin pronunciation and homophone list, and the export function provides the function of exporting homophone vocabulary as a document or table, which is convenient for the user to further use.

[0019] Furthermore, the pronunciation feature extraction unit extracts pronunciation features from the converted text or original speech, specifically in the following steps:

[0020] Text cleaning: remove irrelevant characters in the text, such as punctuation marks, numbers, etc., and only retain Chinese characters or pinyin representations;

[0021] Unvoiced / voiced identification: Determine whether each syllable is unvoiced or voiced through speech signal processing or text-based rules, and perform spectral analysis on the speech signal;

[0022] Final vowel extraction: extract the final vowel of each Chinese character according to the pinyin of the Chinese character or the spectrum characteristics of the speech signal. The final vowel is the part of the Chinese character pronunciation excluding the initial consonant.

[0023] Swallowed / long sound detection: Determine whether there is swallowed or long sound phenomenon through the duration, energy and other characteristics of the speech signal. Swallowed sound is manifested as a significant shortening of the duration of a syllable, while long sound is manifested as a prolongation of the duration.

[0024] Distinguish between flat tongue sounds and curled tongue sounds: For the tip-of-the-tongue front sounds (flat tongue sounds) and tip-of-the-tongue back sounds (curled tongue sounds) in Chinese, the spectral characteristics of the speech signal and the tongue position during pronunciation are used to determine;

[0025] Feature representation: The extracted features are represented in a structured manner, including using vectors or matrices, to facilitate subsequent analysis and comparison.

[0026] Furthermore, the pronunciation comparison and analysis unit receives standard pronunciation feature data from the vocabulary library module, and compares and analyzes the extracted pronunciation features with the standard pronunciation features, specifically in the following steps:

[0027] Feature comparison: compare the pronunciation features to be analyzed with the features in the vocabulary module, and calculate the similarity or distance between the features, such as using indicators such as Euclidean distance and cosine similarity;

[0028] Pronunciation accuracy assessment: Based on the result of feature comparison, the accuracy of the pronunciation to be analyzed is assessed. If the feature similarity is high, the pronunciation is accurate; if the similarity is low, the pronunciation is biased;

[0029] Pronunciation characteristic analysis: Analyze the characteristics of the pronunciation to be analyzed, such as whether there are specific pronunciation habits or accents.

[0030] Furthermore, the grammar analysis unit performs grammar analysis on homophones using the rules in the grammar rule base, specifically in the following steps:

[0031] Homophone matching:

[0032] The syntax analysis unit receives the matching result provided by the homophone matching module;

[0033] Part-of-speech tagging: Tag each matched homophone to determine its possible part of speech in the sentence (such as noun, verb, adjective, etc.);

[0034] Sentence component analysis: According to the syntactic rules of Shaanxi dialect provided in the grammar rule library unit, analyze the possible components (such as subject, predicate, object, etc.) of each homophone in the sentence;

[0035] Collocation check: Check whether the collocation between homophones is reasonable according to the semantic rules of Shaanxi dialect;

[0036] Grammar check: Comprehensively consider the part of speech, sentence components and collocation relationship, and conduct overall grammar check on the generated homophones. If grammatical errors are found, such as mismatched part of speech, incomplete sentence components or improper collocation, the wrong homophones will be corrected or excluded;

[0037] Optimize output: According to the results of grammar verification, optimize homophones to ensure that the generated homophones are grammatically correct and conform to the expression habits of Shaanxi dialect.

[0038] The present invention has the following beneficial effects:

[0039] 1. In the present invention, through the collaborative work of the pronunciation analysis module, the homophone matching module and the grammatical analysis module, the system can accurately generate homophone words that conform to the grammatical rules of the Shaanxi dialect, thereby improving the accuracy and reliability of the results.

[0040] 2. The present invention can automatically analyze the pronunciation characteristics of Shaanxi dialect and quickly generate homophones, which greatly improves the efficiency and accuracy of homophone search, helps to protect and inherit the local cultural heritage of Shaanxi dialect, and also provides a powerful tool for dialect learning and education. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The present invention is a system block diagram of a system for automatically generating dialect homophone words. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] See also Figure 1 As shown, the present invention is a system for automatically generating dialect homophone words, comprising:

[0044] Pronunciation analysis module: responsible for analyzing the pronunciation characteristics of Shaanxi dialect, including voiceless sounds, finals, swallowed sounds, long sounds, flat tongue sounds and curled tongue sounds;

[0045] Vocabulary database module: responsible for storing Shaanxi dialect vocabulary, including ancient vocabulary and unique vocabulary. By establishing a vocabulary database, a large number of Shaanxi dialect vocabulary are collected, classified and annotated. At the same time, users are allowed to add and edit vocabulary;

[0046] Homophone matching module: responsible for searching for words with the same pronunciation in the vocabulary based on the input pronunciation or text;

[0047] Syntax analysis module: responsible for analyzing the grammatical structure of Shaanxi dialect and ensuring that the generated homophones conform to the grammatical rules of Shaanxi dialect;

[0048] User interface module: responsible for providing user input and output interfaces, including voice input, text input and display of homophones;

[0049] The pronunciation analysis module needs to receive the voice or text input by the user and perform pronunciation analysis on it. The pronunciation analysis module outputs the pronunciation features obtained by the analysis to the homophone matching module, and updates the new pronunciation data or rules to the vocabulary library module. The homophone matching module accesses the vocabulary database in the vocabulary library module to obtain vocabulary information in the Shaanxi dialect. The homophone matching module searches for words with the same pronunciation in the vocabulary database based on the pronunciation features provided by the pronunciation analysis module, and outputs the matching results to the grammar analysis module. The grammar analysis module receives the matching results provided by the homophone matching module, performs grammar analysis and verification on the matching results, and outputs the verified homophone vocabulary to the user interface module.

[0050] The working principle of the system for automatically generating dialect homophone vocabulary proposed by the present invention is that the user inputs voice or text through the user interface module, the user interface module transmits the input to the pronunciation analysis module, the pronunciation analysis module performs pronunciation analysis on the input and accesses the pronunciation database in the vocabulary library module, the pronunciation analysis module outputs the pronunciation features obtained by the analysis to the homophone matching module, the homophone matching module searches for words with the same pronunciation in the vocabulary database in the vocabulary library module and matches them, the homophone matching module outputs the matching results to the grammar analysis module, the grammar analysis module performs grammar analysis and verification on the matching results, the grammar analysis module outputs the verified homophone vocabulary to the user interface module, the user interface module displays the homophone vocabulary to the user and may receive feedback or new input from the user.

[0051] In one embodiment, for the above-mentioned pronunciation analysis module, the pronunciation analysis module includes a speech recognition unit, a pronunciation feature extraction unit, a pronunciation comparison analysis unit and a result output unit. The speech recognition unit is responsible for converting the input Shaanxi dialect speech into text, the speech recognition unit passes the converted text to the pronunciation feature extraction unit, the pronunciation feature extraction unit is responsible for extracting pronunciation features from the converted text or original speech, the pronunciation feature extraction unit passes the extracted pronunciation feature data to the pronunciation comparison analysis unit, the pronunciation comparison analysis unit is responsible for comparing and analyzing the extracted pronunciation features with the standard pronunciation features, the pronunciation comparison analysis unit receives the standard pronunciation feature data from the vocabulary library module, the pronunciation comparison analysis unit passes the analysis result to the result output unit, and the result output unit is responsible for outputting the result of the pronunciation analysis to the homophone matching module.

[0052] In one embodiment, for the above-mentioned vocabulary library module, the vocabulary library module includes a database unit, a user interface unit, a business logic unit, a data verification unit and a log recording unit. The database unit is responsible for storing the data of Shaanxi dialect vocabulary, including vocabulary ID, classification, pronunciation, explanation, usage examples and other fields. The database unit provides operation interfaces such as data query, insertion, update and deletion. The user interface unit is responsible for receiving user input, such as adding new vocabulary, editing existing vocabulary, etc., and passing the user request to the database unit for processing. The business logic unit is responsible for processing the business logic of the user request, such as verifying the validity of the user input, generating a new vocabulary ID, etc. The data verification unit is responsible for verifying the data input by the user to ensure that the data is correct. The legitimacy and accuracy of the data, such as checking whether the vocabulary classification entered by the user is valid and whether the pronunciation conforms to the specifications, etc. The log recording unit records the user operation log and the system operation status log to facilitate problem troubleshooting and data analysis, and provides a log query function to allow users to view historical operation records. The user interface unit receives the user input and passes it to the business logic unit for processing. After the business logic unit processes the user request, it returns the result to the user interface unit. The business logic unit calls the data operation interface provided by the database unit according to the user request. Before processing the user request, the business logic unit calls the data verification unit to verify the user input. When the log recording unit receives event notifications from other units, it records the corresponding log information.

[0053] In one embodiment, for the above-mentioned homophone matching module, the homophone matching module uses the pronunciation features provided by the pronunciation analysis module as query conditions, calls the data query interface of the vocabulary library module, and the vocabulary library module returns a vocabulary list that matches the input pronunciation according to the query conditions, filters the returned vocabulary list, removes vocabulary that does not meet the conditions (such as vocabulary type mismatch, incorrect pronunciation annotation, etc.), sorts the filtered vocabulary, and generates a final matching result list.

[0054] In one embodiment, for the above-mentioned grammatical analysis module, the grammatical analysis module includes a grammatical rule base unit, a grammatical analysis unit, a grammatical verification unit, and a result optimization and feedback unit. The grammatical rule base unit stores the grammatical rules of Shaanxi dialect, including lexical, syntactic and semantic rules. The grammatical rule base unit provides a rule query interface, allowing other units to perform rule query as needed. The grammatical analysis unit uses the rules in the grammatical rule base to perform grammatical analysis on homophones. The grammatical analysis unit checks whether the vocabulary collocation and sentence structure in the homophones conform to the grammatical rules of Shaanxi dialect. The grammatical verification unit performs overall grammatical verification on the homophones to ensure that the entire sentence or vocabulary combination is grammatically correct. The grammatical verification unit provides error prompts and suggestions to help users make corrections. The result optimization and feedback unit optimizes the homophones according to the grammatical analysis results. The grammatical analysis unit receives the homophones generated by the homophone matching module, and performs grammatical analysis and verification on the homophones.

[0055] In one embodiment, for the above-mentioned user interface module, the user interface module includes an input unit and an output unit. The input unit includes voice input and text input. The voice input is the user inputting the voice of Shaanxi dialect through the microphone of a mobile phone, computer and other devices. The text input is the user manually inputting the text of Shaanxi dialect, which is convenient for homophone search for specific words. The output unit includes homophone vocabulary display and export function. The homophone vocabulary display displays the matched homophones to the user in the form of a list, including dialect vocabulary, Mandarin pronunciation and homophone list. The export function provides the function of exporting homophone vocabulary as a document or table, which is convenient for the user to further use.

[0056] In one embodiment, for the above-mentioned pronunciation feature extraction unit, the pronunciation feature extraction unit extracts pronunciation features from the converted text or original speech, specifically in the following steps:

[0057] Text cleaning: remove irrelevant characters in the text, such as punctuation marks, numbers, etc., and only retain Chinese characters or pinyin representations;

[0058] Unvoiced / voiced identification: Determine whether each syllable is unvoiced or voiced through speech signal processing or text-based rules, and perform spectral analysis on the speech signal;

[0059] Final vowel extraction: Extract the final vowel of each Chinese character based on the pinyin of the Chinese character or the spectral characteristics of the speech signal. The final vowel is the part of the Chinese character pronunciation other than the initial consonant.

[0060] Swallowed / long sound detection: Determine whether there is swallowed or long sound phenomenon through the duration, energy and other characteristics of the speech signal. Swallowed sound is manifested as a significant shortening of the duration of a syllable, while long sound is manifested as a prolongation of the duration.

[0061] Distinguish between flat tongue sounds and curled tongue sounds: For the tip-of-the-tongue front sounds (flat tongue sounds) and tip-of-the-tongue back sounds (curled tongue sounds) in Chinese, the spectral characteristics of the speech signal and the tongue position during pronunciation are used to determine;

[0062] Feature representation: The extracted features are represented in a structured manner, including using vectors or matrices, to facilitate subsequent analysis and comparison.

[0063] In one embodiment, for the above-mentioned pronunciation comparison and analysis unit, the pronunciation comparison and analysis unit receives the standard pronunciation feature data from the vocabulary library module, and compares and analyzes the extracted pronunciation features with the standard pronunciation features. The specific steps are:

[0064] Feature comparison: compare the pronunciation features to be analyzed with the features in the vocabulary module, and calculate the similarity or distance between the features, such as using indicators such as Euclidean distance and cosine similarity;

[0065] Pronunciation accuracy assessment: Based on the result of feature comparison, the accuracy of the pronunciation to be analyzed is assessed. If the feature similarity is high, the pronunciation is accurate; if the similarity is low, the pronunciation is biased;

[0066] Pronunciation characteristics analysis: Analyze the characteristics of the pronunciation to be analyzed, such as whether there are specific pronunciation habits or accents;

[0067] Suppose we have a Shaanxi dialect pronunciation to be analyzed, “shàng xué” (go to school);

[0068] Feature extraction: For "shàng", the final "àng" is extracted and judged as a retroflex consonant and a long consonant; for "xué", the final "ué" is extracted and judged as a voiceless consonant;

[0069] Feature comparison: compare the extracted features with the pronunciation features of "shàng xué" in the standard database, and calculate the similarity between the features. For example, using cosine similarity, 0.95 is obtained (assumed value, for illustration only);

[0070] Pronunciation accuracy assessment: Due to the high similarity (0.95), it can be judged that the pronunciation of "shàng xué" to be analyzed is relatively accurate;

[0071] Analysis of pronunciation characteristics: If during the comparison process, it is found that some features are significantly different from those in the vocabulary module (such as the pronunciation duration of the finals is significantly different), further analysis is conducted to determine whether these differences reflect specific pronunciation habits or accents.

[0072] In one embodiment, for the above-mentioned syntax analysis unit, the syntax analysis unit performs syntax analysis on homophones using rules in a syntax rule base, specifically in the following steps:

[0073] Homophone matching:

[0074] The syntactic analysis unit receives the matching results provided by the homophone matching module;

[0075] Part-of-speech tagging: Perform part-of-speech tagging on each matched homophone to determine its possible part of speech in the sentence (such as noun, verb, adjective, etc.);

[0076] Sentence component analysis: According to the syntactic rules of Shaanxi dialect provided in the grammar rule library unit, analyze the possible components that each homophone can act as in the sentence (such as subject, predicate, object, etc.);

[0077] Collocation relationship check: According to the semantic rules of Shaanxi dialect, check whether the collocation relationship between homophones is reasonable;

[0078] Grammar verification: Considering the part of speech, sentence components, and collocation relationship comprehensively, perform overall grammar verification on the generated homophone vocabulary. If grammar errors are found, such as mismatched parts of speech, incomplete sentence components, or improper collocations, correct or exclude the incorrect homophones;

[0079] Optimized output: According to the results of grammar verification, optimize the homophone vocabulary to ensure that the generated homophone vocabulary is grammatically correct and conforms to the expression habits of Shaanxi dialect;

[0080] Suppose the text input by the user is "我吃了一个苹果" (I ate an apple), and the system matches the homophones "我" and "鹅" (which may have similar pronunciations in Shaanxi dialect, but this is just an example here);

[0081] Part-of-speech tagging:

[0082] "我" is tagged as a pronoun;

[0083] "鹅" is tagged as a noun, but it is obviously not suitable in this context;

[0084] Sentence component analysis:

[0085] "我" serves as the subject of the sentence;

[0086] If "鹅" is used as the subject, it does not match the subsequent "吃了一个苹果";

[0087] Collocation relationship check:

[0088] "我吃了一个苹果" is a reasonable collocation relationship;

[0089] "鹅吃了一个苹果" may be grammatically correct, but it does not make sense semantically;

[0090] Grammar verification and optimization:

[0091] Exclude the option of "goose" as a homophone;

[0092] Confirm that "I" is the correct homophone and generate homophonic vocabulary that conforms to the grammar rules of Shaanxi dialect (in this example, there may be only one "I" that meets the conditions).

[0093] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A system for automatically generating dialect homophones, characterized in that: include: Pronunciation analysis module: responsible for analyzing the pronunciation characteristics of Shaanxi dialect; Vocabulary library module: responsible for storing vocabulary in Shaanxi dialect, including ancient vocabulary and unique vocabulary; Homophone matching module: responsible for searching for words with the same pronunciation in the vocabulary based on the input pronunciation or text; Syntax analysis module: responsible for analyzing the grammatical structure of Shaanxi dialect and ensuring that the generated homophones conform to the grammatical rules of Shaanxi dialect; User interface module: responsible for providing user input and output interfaces, including voice input, text input and display of homophones; The pronunciation analysis module needs to receive the voice or text input by the user and perform pronunciation analysis on it. The pronunciation analysis module outputs the pronunciation features obtained by the analysis to the homophone matching module, and updates the new pronunciation data or rules to the vocabulary library module. The homophone matching module accesses the vocabulary database in the vocabulary library module to obtain vocabulary information in the Shaanxi dialect. The homophone matching module searches for words with the same pronunciation in the vocabulary database according to the pronunciation features provided by the pronunciation analysis module, and outputs the matching results to the grammar analysis module. The grammar analysis module receives the matching results provided by the homophone matching module, performs grammar analysis and verification on the matching results, and outputs the verified homophone vocabulary to the user interface module.

2. A system for automatically generating homophones in dialects according to claim 1, characterized in that: The pronunciation analysis module includes a speech recognition unit, a pronunciation feature extraction unit, a pronunciation comparison analysis unit and a result output unit. The speech recognition unit is responsible for converting the input Shaanxi dialect speech into text, the speech recognition unit passes the converted text to the pronunciation feature extraction unit, the pronunciation feature extraction unit is responsible for extracting pronunciation features from the converted text or original speech, the pronunciation feature extraction unit passes the extracted pronunciation feature data to the pronunciation comparison analysis unit, the pronunciation comparison analysis unit is responsible for comparing and analyzing the extracted pronunciation features with standard pronunciation features, the pronunciation comparison analysis unit receives standard pronunciation feature data from the vocabulary library module, the pronunciation comparison analysis unit passes the analysis result to the result output unit, and the result output unit is responsible for outputting the result of the pronunciation analysis to the homophone matching module.

3. The system for automatically generating homophones in dialects according to claim 1, characterized in that: The vocabulary library module includes a database unit, a user interface unit, a business logic unit, a data verification unit and a log recording unit. The database unit is responsible for storing data of Shaanxi dialect vocabulary. The database unit provides operation interfaces such as data query, insertion, update and deletion. The user interface unit is responsible for receiving user input and passing the user request to the database unit for processing. The business logic unit is responsible for processing the business logic of the user request. The data verification unit is responsible for verifying the data input by the user. The log recording unit records the user operation log and the system operation status log. The user interface unit receives the user input and passes it to the business logic unit for processing. After processing the user request, the business logic unit returns the result to the user interface unit. The business logic unit calls the data operation interface provided by the database unit according to the user request. Before processing the user request, the business logic unit calls the data verification unit to verify the user input. When the log recording unit receives event notifications from other units, it records corresponding log information.

4. The system for automatically generating homophones in dialects according to claim 1, characterized in that: The homophone matching module uses the pronunciation features provided by the pronunciation analysis module as query conditions and calls the data query interface of the vocabulary library module. The vocabulary library module returns a vocabulary list that matches the input pronunciation according to the query conditions, filters the returned vocabulary list, removes words that do not meet the conditions, sorts the filtered words, and generates a final matching result list.

5. The system for automatically generating homophones in dialects according to claim 1, characterized in that: The grammatical analysis module includes a grammatical rule base unit, a grammatical analysis unit, a grammatical verification unit and a result optimization and feedback unit. The grammatical rule base unit stores the grammatical rules of Shaanxi dialect, including lexical, syntactic and semantic rules. The grammatical analysis unit uses the rules in the grammatical rule base to perform grammatical analysis on homophones. The grammatical analysis unit checks whether the vocabulary collocation and sentence structure in the homophones conform to the grammatical rules of Shaanxi dialect. The grammatical verification unit performs overall grammatical verification on the homophones. The result optimization and feedback unit optimizes the homophones according to the grammatical analysis results. The grammatical analysis unit receives the homophones generated by the homophone matching module, and performs grammatical analysis and verification on the homophones.

6. The system for automatically generating homophones in dialects according to claim 1, characterized in that: The user interface module includes an input unit and an output unit. The input unit includes voice input and text input. The voice input is the voice of the Shaanxi dialect input by the user through a microphone. The text input is the text of the Shaanxi dialect input manually by the user. The output unit includes homophone vocabulary display and export functions. The homophone vocabulary display displays the matched homophones to the user in the form of a list, including dialect vocabulary, Mandarin pronunciation and a homophone list. The export function provides a function of exporting homophone vocabulary as a document or a table.

7. The system for automatically generating homophones in dialects according to claim 2, characterized in that: The pronunciation feature extraction unit extracts pronunciation features from the converted text or original speech, specifically in the following steps: Text cleaning: remove irrelevant characters from the text and keep only Chinese characters or pinyin representations; Unvoiced / voiced identification: Determine whether each syllable is unvoiced or voiced through speech signal processing or text-based rules; Final vowel extraction: Extract the final vowel of each Chinese character based on the pinyin of the Chinese character or the spectral characteristics of the speech signal; Swallowed / long sound detection: Determine whether there is swallowed or long sound phenomenon based on the characteristics of the speech signal; Distinguish between flat tongue consonants and curled tongue consonants: For the tip-of-the-tongue front consonants and tip-of-the-tongue back consonants in Chinese, the spectral characteristics of the speech signal and the tongue position during pronunciation are used to determine; Feature representation: The extracted features are represented in a structured manner.

8. The system for automatically generating homophones in dialects according to claim 2, characterized in that: The pronunciation comparison and analysis unit receives standard pronunciation feature data from the vocabulary library module and compares and analyzes the extracted pronunciation features with the standard pronunciation features. The specific steps are as follows: Feature comparison: compare the pronunciation features to be analyzed with the features in the vocabulary module and calculate the similarity or distance between the features; Pronunciation accuracy assessment: Based on the result of feature comparison, the accuracy of the pronunciation to be analyzed is assessed. If the feature similarity is high, the pronunciation is accurate; if the similarity is low, the pronunciation is biased; Pronunciation characteristic analysis: Analyze the characteristics of the pronunciation to be analyzed, such as whether there are specific pronunciation habits or accents.

9. The system for automatically generating homophones in dialects according to claim 5, characterized in that: The grammar analysis unit performs grammar analysis on homophones using the rules in the grammar rule base, specifically in the following steps: Homophone matching: The syntax analysis unit receives the matching result provided by the homophone matching module; Part-of-speech tagging: Tag each matched homophone to determine its possible part of speech in the sentence; Sentence component analysis: According to the grammatical rules of Shaanxi dialect provided in the grammar rule library unit, analyze the possible components of each homophone in the sentence; Collocation check: Check whether the collocation between homophones is reasonable according to the semantic rules of Shaanxi dialect; Grammar check: Comprehensively consider the part of speech, sentence components and collocation relationship, and conduct overall grammar check on the generated homophones. If grammatical errors are found, such as mismatched part of speech, incomplete sentence components or improper collocation, the wrong homophones will be corrected or excluded; Optimize output: Optimize homophones based on the results of grammar verification.

Images