Chinese-English bidirectional machine translation test method based on grammar composition analysis

Abstract

The invention discloses a Chinese-English two-way machine translation test method based on grammar composition analysis, and belongs to the technical field of Chinese-English translation, and the method comprises the following steps: S1, collecting a Chinese-English bilingual corpus, processing the Chinese-English bilingual corpus, and determining a Chinese-English bilingual corpus association set; s2, constructing a Chinese-English two-way machine translation test model, and determining an optimal Chinese-English two-way machine translation test model; and S3, translating the machine translation demand data based on the optimal Chinese-English two-way machine translation test model, and determining a Chinese-English two-way machine translation test result. According to the method, the problems that the Chinese and English bidirectional machine translation accuracy is low and the Chinese and English bidirectional machine translation quality is reduced due to the fact that the Chinese and English bidirectional translation cannot be performed based on grammar composition analysis in the prior art are solved. According to the method, Chinese and English bidirectional translation can be performed based on grammatical composition analysis, so that the Chinese-English bidirectional machine translation accuracy is high, and the Chinese-English bidirectional machine translation quality can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Chinese-English translation, and in particular to a Chinese-English bidirectional machine translation testing method based on grammatical composition analysis. Background Art

[0002] With the rapid development of Internet technology and the accelerated advancement of the globalization process, more and more international exchanges and cooperation have become more frequent; in this case, translation, as an important bridge connecting communication between different languages, is of self-evident importance.

[0003] The Chinese patent with publication number CN210466389U discloses an intelligent Chinese-English voice translator, including: a voice collection system and a voice translation system, the voice collection system includes a voice input module and a voice recognition module, the output end of the voice input module is connected to the input end of the voice recognition module; the voice translation system includes a translation controller and a voice playback module, the output end of the voice recognition module is connected to the input end of the translation controller, and the output end of the translation controller is connected to the input end of the voice playback module; the voice translation system can quickly complete Chinese-English simultaneous interpretation, which is suitable for English learning, travel communication, etc.; However, the patent has the following defects:

[0004] The existing bidirectional translation between Chinese and English cannot be performed based on grammatical composition analysis, which makes the accuracy of bidirectional machine translation between Chinese and English low and reduces the quality of bidirectional machine translation between Chinese and English. Summary of the invention

[0005] The purpose of the present invention is to provide a Chinese-English bidirectional machine translation test method based on grammatical composition analysis, which can perform bidirectional translation between Chinese and English based on grammatical composition analysis, so that the accuracy of Chinese-English bidirectional machine translation is high, the quality of Chinese-English bidirectional machine translation can be improved, and the problems raised in the above-mentioned background technology can be solved.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis includes the following steps:

[0008] S1: Collect Chinese and English bilingual corpora, process them, and determine the Chinese and English bilingual corpus association set based on grammatical composition analysis;

[0009] S2: According to the test requirements of Chinese-English bidirectional machine translation, a Chinese-English bidirectional machine translation test model based on grammatical composition analysis is constructed, and the Chinese-English bidirectional machine translation test model is evaluated and optimized to determine the optimal Chinese-English bidirectional machine translation test model;

[0010] S3: Translate the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model, and determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis;

[0011] S4: Based on the Chinese-English bidirectional machine translation test optimization and control solution, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled to meet the needs of Chinese-English bidirectional machine translation.

[0012] Preferably, in S1, a Chinese-English bilingual corpus is collected, and the following operations are performed:

[0013] Acquire Chinese materials in different fields and topics to determine Chinese corpus data;

[0014] Obtain English materials in different fields and topics to determine English corpus data;

[0015] Among them, Chinese-English bilingual corpus data are determined based on Chinese corpus data and English corpus data.

[0016] Preferably, in S1, the Chinese-English bilingual corpus is processed by performing the following operations:

[0017] Obtain bilingual corpus data in Chinese and English;

[0018] Based on the sequential retrieval method, the Chinese and English bilingual corpus data are retrieved;

[0019] Remove inconsistent, invalid and missing data in the Chinese-English bilingual corpus data that are of no value for Chinese-English bidirectional machine translation testing;

[0020] Identify bilingual corpus data that is valuable for Chinese-English bidirectional machine translation testing;

[0021] Obtain Chinese-English bilingual corpus data that is valuable for Chinese-English bidirectional machine translation testing;

[0022] Perform word segmentation, part-of-speech tagging, syntax and grammar analysis on the valuable Chinese-English bilingual corpus data for Chinese-English bidirectional machine translation tests to make the source language and target language semantically matched;

[0023] Determine the associated set of Chinese and English bilingual corpora based on grammatical composition analysis.

[0024] Preferably, in S1, processing the Chinese-English bilingual corpus further includes:

[0025] Extracting the amount of inconsistent data information that is worthless for Chinese-English bidirectional machine translation testing from the English bilingual corpus data as the first data amount;

[0026] Extracting the amount of invalid data information that is worthless to the Chinese-English bidirectional machine translation test from the English bilingual corpus data as the second data amount;

[0027] Extracting the amount of data information of sentences containing missing data that is worthless for Chinese-English bidirectional machine translation testing from the English bilingual corpus data as the third amount of data; wherein the sentence containing the missing data is a sentence with missing words or characters;

[0028] The first abnormality evaluation coefficient is obtained by using the first data amount, the second data amount and the third data amount, wherein the first abnormality evaluation coefficient is obtained by the following formula:

[0029]

[0030] Among them, P 01 Denotes the first abnormal evaluation coefficient; D 01 , D 02 and D 03 Respectively represent the first data volume, the second data volume and the third data volume; α, β and γ represent the first adjustment coefficient, the second adjustment coefficient and the third adjustment coefficient; and the value ranges of the first adjustment coefficient, the second adjustment coefficient and the third adjustment coefficient are: [0.1, 5], [0.3, 2] and [0.2, 1.3];

[0031] Comparing the first abnormality evaluation coefficient with a preset first coefficient threshold;

[0032] When the first abnormal evaluation coefficient exceeds a preset first coefficient threshold, it is determined that the English bilingual corpus data has data quality abnormality, and a data quality abnormality alarm is issued;

[0033] When the first abnormality evaluation coefficient does not exceed the preset first coefficient threshold, the semantic weight value of the sentence containing inconsistent, invalid and missing data is retrieved, and the data quality is secondary checked based on the semantic weight value of the sentence containing inconsistent, invalid and missing data.

[0034] Preferably, when the first abnormal evaluation coefficient does not exceed the preset first coefficient threshold, the semantic weight value of the sentence where the inconsistent, invalid and missing data is located is retrieved, and the data quality is secondary detected according to the semantic weight value of the sentence where the inconsistent, invalid and missing data is located, including:

[0035] Extracting the semantic weight value of the sentence where each inconsistent data is located, wherein the inconsistent data includes Chinese and English sentences that do not conform to the language features;

[0036] Extract the semantic weight value of each sentence containing invalid data;

[0037] Extracting the semantic weight value of each sentence where the missing data is located; wherein the sentence where the missing data is located is a sentence where words or characters are missing;

[0038] The second abnormality evaluation coefficient is obtained by using the semantic weight value of the sentence where each inconsistent data is located, the semantic weight value of the sentence where each invalid data is located, and the semantic weight value of the sentence where each missing data is located; wherein the second abnormality evaluation coefficient is obtained by the following formula:

[0039]

[0040] in

[0041]

[0042] Among them, P 01 Represents the second abnormality evaluation coefficient; N 01 、N 02 and N 03 Respectively represent the number of statements with inconsistent, invalid and missing data; W 01i , W 02i and W 03i Respectively represent the semantic weight values ​​corresponding to the sentences containing the i-th inconsistent, invalid and missing data; W 01h , W 02h and W 03h Respectively represent the semantic weights and values ​​of the sentences containing inconsistent, invalid and missing data; k 01 , k 02 and k 03 represent the first adjustment factor, the second adjustment factor and the third adjustment factor respectively, and the value ranges of the first adjustment factor, the second adjustment factor and the third adjustment factor are: [0.3, 2], [0.6, 3] and [0.8, 1.7];

[0043] comparing the second abnormality evaluation coefficient with a preset second coefficient threshold;

[0044] When the second abnormality evaluation coefficient exceeds a preset second coefficient threshold, it is determined that the English bilingual corpus data has data quality abnormality, and a data quality abnormality alarm is issued.

[0045] Preferably, in S2, a Chinese-English bidirectional machine translation test model based on grammatical composition analysis is constructed, and the following operations are performed:

[0046] Obtain a set of Chinese and English bilingual corpora based on grammatical composition analysis;

[0047] Divide the Chinese-English bilingual corpus association set based on grammatical composition analysis to determine the Chinese-English bidirectional machine translation training set and the Chinese-English bidirectional machine translation test set;

[0048] According to the requirements of Chinese-English bidirectional machine translation test, select a neural network framework suitable for Chinese-English bidirectional machine translation test;

[0049] Based on the Chinese-English bidirectional machine translation training set, train the selected neural network framework suitable for Chinese-English bidirectional machine translation testing;

[0050] A Chinese-English bidirectional machine translation test model based on grammatical composition analysis was determined.

[0051] Preferably, in S2, the Chinese-English bidirectional machine translation test model is evaluated and optimized by performing the following operations:

[0052] Obtain a Chinese-English bidirectional machine translation test model based on grammatical composition analysis;

[0053] Based on the Chinese-English bidirectional machine translation test set, the Chinese-English bidirectional machine translation test model based on grammatical composition analysis was evaluated and tested;

[0054] Determine the evaluation test results based on the Chinese-English bidirectional machine translation test model;

[0055] Conduct mining and analysis on the evaluation test results based on the Chinese-English bidirectional machine translation test model;

[0056] Determine the optimization adjustment plan based on the Chinese-English bidirectional machine translation test model;

[0057] Optimize and adjust the Chinese-English bidirectional machine translation test model according to the optimization adjustment plan based on the Chinese-English bidirectional machine translation test model;

[0058] Determine the optimal Chinese-English bidirectional machine translation test model.

[0059] Preferably, in S3, the machine translation demand data is translated based on the optimal Chinese-English bidirectional machine translation test model, and the following operations are performed:

[0060] According to the requirements of Chinese-English bidirectional machine translation testing, obtain machine translation demand data;

[0061] Input machine translation demand data into the optimal Chinese-English bidirectional machine translation test model;

[0062] Translate the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model;

[0063] Determine the Chinese-English bidirectional machine translation results based on grammatical composition analysis;

[0064] Compare and analyze the Chinese-English bidirectional machine translation results with the preset Chinese-English bidirectional translation standards;

[0065] Determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis.

[0066] Preferably, in S3, the Chinese-English bidirectional machine translation test results based on grammatical composition analysis include:

[0067] When the Chinese-English bidirectional machine translation result matches the preset Chinese-English bidirectional translation standard, the Chinese-English bidirectional machine translation test result based on grammatical composition analysis is that the Chinese-English bidirectional machine translation is normal;

[0068] When the Chinese-English bidirectional machine translation result does not match the preset Chinese-English bidirectional translation standard, the Chinese-English bidirectional machine translation test result based on grammatical composition analysis is that the Chinese-English bidirectional machine translation is abnormal.

[0069] Preferably, in S4, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled, and the following operations are performed:

[0070] Obtain Chinese-English bidirectional machine translation test results based on grammatical composition analysis;

[0071] Conduct in-depth mining and analysis of Chinese-English bidirectional machine translation test results based on grammatical composition analysis;

[0072] Determine the optimization and control plan for Chinese-English bidirectional machine translation testing based on grammatical composition analysis;

[0073] Based on the Chinese-English bidirectional machine translation test optimization and control solution, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled;

[0074] The model structure and parameters of the optimal Chinese-English bidirectional machine translation test model are continuously adjusted, and multiple iterations are performed to determine a Chinese-English bidirectional machine translation test model that can meet the needs of Chinese-English bidirectional machine translation.

[0075] Compared with the prior art, the present invention has the following beneficial effects:

[0076] The present invention collects Chinese-English bilingual corpora, processes the Chinese-English bilingual corpora, determines a Chinese-English bilingual corpus association set based on grammatical composition analysis, constructs a Chinese-English bilingual machine translation test model based on grammatical composition analysis according to Chinese-English bilingual machine translation test requirements, evaluates and optimizes the Chinese-English bilingual machine translation test model, determines an optimal Chinese-English bilingual machine translation test model, translates machine translation demand data based on the optimal Chinese-English bilingual machine translation test model, determines a Chinese-English bilingual machine translation test result based on grammatical composition analysis, formulates a Chinese-English bilingual machine translation test optimization management and control plan based on the Chinese-English bilingual machine translation test result, performs intelligent optimization and management on the optimal Chinese-English bilingual machine translation test model based on the Chinese-English bilingual machine translation test optimization management and control plan, meets Chinese-English bilingual machine translation requirements, can perform bidirectional translation between Chinese and English based on grammatical composition analysis, makes Chinese-English bilingual machine translation highly accurate, and improves Chinese-English bilingual machine translation quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 The figure is a flow chart of the Chinese-English bidirectional machine translation testing method based on grammatical composition analysis of the present invention. DETAILED DESCRIPTION

[0078] 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.

[0079] In order to solve the problem that the existing bidirectional translation between Chinese and English cannot be performed based on grammatical composition analysis, which makes the accuracy of bidirectional machine translation between Chinese and English low and reduces the quality of bidirectional machine translation between Chinese and English, please refer to Figure 1 , this embodiment provides the following technical solutions:

[0080] The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis includes the following steps:

[0081] S1: Collect Chinese and English bilingual corpora, process them, and determine the Chinese and English bilingual corpus association set based on grammatical composition analysis;

[0082] In this embodiment, a Chinese-English bilingual corpus is collected and the following operations are performed:

[0083] Acquire Chinese materials in different fields and topics to determine Chinese corpus data;

[0084] Obtain English materials in different fields and topics to determine English corpus data;

[0085] Among them, Chinese-English bilingual corpus data are determined based on Chinese corpus data and English corpus data.

[0086] It should be noted that Chinese and English materials include various types of texts such as articles, books, news reports, forum posts, and social media comments;

[0087] By acquiring a large amount of Chinese and English corpus data, Chinese-English bilingual corpus data can be determined to provide data support for Chinese-English two-way machine translation.

[0088] In this embodiment, the Chinese-English bilingual corpus is processed by performing the following operations:

[0089] Obtain bilingual corpus data in Chinese and English;

[0090] Based on the sequential retrieval method, the Chinese and English bilingual corpus data are retrieved;

[0091] Remove inconsistent, invalid and missing data in the Chinese-English bilingual corpus data that are of no value for Chinese-English bidirectional machine translation testing;

[0092] Identify bilingual corpus data that is valuable for Chinese-English bidirectional machine translation testing;

[0093] It should be noted that when searching for Chinese-English bilingual corpus data, firstly, the consistency check is performed on the Chinese-English bilingual corpus data. According to the reasonable value range and mutual relationship of the Chinese-English bilingual corpus data, the Chinese-English bilingual corpus data is checked to see whether it meets the requirements, and inconsistent data that exceeds the normal range, is logically unreasonable or contradictory in the Chinese-English bilingual corpus data is removed. Secondly, the invalid value and missing value processing is performed on the Chinese-English bilingual corpus data, and the invalid data and missing data that are of no value to the Chinese-English bilingual machine translation test are removed. The Chinese-English bilingual corpus data that are valuable to the Chinese-English bilingual machine translation test can be determined, so as to facilitate the subsequent word segmentation, part-of-speech tagging, syntactic and grammatical analysis of the Chinese-English bilingual corpus data that are valuable to the Chinese-English bilingual machine translation test.

[0094] Obtain Chinese-English bilingual corpus data that is valuable for Chinese-English bidirectional machine translation testing;

[0095] Perform word segmentation, part-of-speech tagging, syntax and grammar analysis on the valuable Chinese-English bilingual corpus data for Chinese-English bidirectional machine translation tests to make the source language and target language semantically matched;

[0096] Determine the associated set of Chinese and English bilingual corpora based on grammatical composition analysis.

[0097] It should be noted that word segmentation is the process of recombining continuous character sequences into semantically independent word sequences according to certain specifications. In English texts, spaces are used as natural delimiters between words, while in Chinese, characters, sentences and paragraphs can only be simply demarcated by obvious delimiters. Among them, the combination rules of vocabulary are learned by training bilingual corpus data in Chinese and English, and then these rules are applied to new texts for word segmentation. Word segmentation is the basis of text mining. For a piece of input text, successful word segmentation can achieve the effect of automatic computer recognition of the meaning of the sentence.

[0098] It should be noted that part-of-speech tagging, also known as grammatical tagging or word class resolution, is a text data processing technology in corpus linguistics that marks the parts of speech of words in a corpus according to their meaning and contextual content. Part-of-speech tagging can be done manually or by specific algorithms. Using machine learning methods to implement part-of-speech tagging is the research content of natural language processing. Common part-of-speech tagging algorithms include hidden Markov models, conditional random fields, etc. After part-of-speech tagging, the accuracy and efficiency of subsequent natural language processing tasks can be improved, such as syntactic analysis, semantic analysis, etc.

[0099] It should be noted that syntactic analysis refers to the analysis of the grammatical functions of words in a sentence. For example, in "I came late", "I" is the subject, "come" is the predicate, and "late" is the complement. The main application of syntactic analysis is in information processing, such as machine translation. It is a direct implementation of the idea of ​​chunk analysis. Chunk analysis simplifies the description of sentences by identifying high-level structural units. One way to find chunk rules from different sentences is to learn a grammar that can explain the chunk structure we have found.

[0100] It should be noted that grammatical analysis is a logical stage of the compilation process. The task of grammatical analysis is to combine word sequences into various grammatical phrases, such as "program", "statement", "expression", etc. on the basis of lexical analysis. The program that completes the grammatical analysis task is called a grammatical analyzer, or a grammatical analysis program. According to the grammatical rules of the source language, the corresponding grammatical categories are identified from the results of lexical analysis, and grammar checking is performed at the same time. According to the grammatical rules of the source language, the corresponding grammatical categories are identified from the results of lexical analysis, and grammar checking is performed at the same time.

[0101] Specifically, in S1, the Chinese-English bilingual corpus is processed, and further includes:

[0102] Extracting the amount of inconsistent data information that is worthless for Chinese-English bidirectional machine translation testing from the English bilingual corpus data as the first data amount;

[0103] Extracting the amount of invalid data information that is worthless to the Chinese-English bidirectional machine translation test from the English bilingual corpus data as the second data amount;

[0104] Extracting the amount of data information of sentences containing missing data that is worthless for Chinese-English bidirectional machine translation testing from the English bilingual corpus data as the third amount of data; wherein the sentence containing the missing data is a sentence with missing words or characters;

[0105] The first abnormality evaluation coefficient is obtained by using the first data amount, the second data amount and the third data amount, wherein the first abnormality evaluation coefficient is obtained by the following formula:

[0106]

[0107] Among them, P 01 Denotes the first abnormal evaluation coefficient; D 01 , D 02 and D 03 Respectively represent the first data volume, the second data volume and the third data volume; α, β and γ represent the first adjustment coefficient, the second adjustment coefficient and the third adjustment coefficient; and the value ranges of the first adjustment coefficient, the second adjustment coefficient and the third adjustment coefficient are: [0.1, 5], [0.3, 2] and [0.2, 1.3];

[0108] Comparing the first abnormality evaluation coefficient with a preset first coefficient threshold;

[0109] When the first abnormal evaluation coefficient exceeds a preset first coefficient threshold, it is determined that the English bilingual corpus data has data quality abnormality, and a data quality abnormality alarm is issued;

[0110] When the first abnormality evaluation coefficient does not exceed the preset first coefficient threshold, the semantic weight value of the sentence containing inconsistent, invalid and missing data is retrieved, and the data quality is secondary checked based on the semantic weight value of the sentence containing inconsistent, invalid and missing data.

[0111] The technical effect of the above technical solution is: by extracting and analyzing the amount of inconsistent, invalid and missing data in the Chinese and English bilingual corpus, the technical solution can comprehensively evaluate the quality of the data. This helps to identify data problems that may have a negative impact on machine translation tests. Using the first abnormal evaluation coefficient, the technical solution can quantify the severity of data quality problems and compare them with the preset threshold. When the data quality is abnormal, the system will issue an alarm, which helps to timely discover and deal with data problems and ensure the accuracy and reliability of machine translation tests. For cases that do not exceed the first coefficient threshold, the technical solution will further retrieve the semantic weight value of the sentence where the inconsistent, invalid and missing data are located for secondary detection. This refined data processing method can more accurately identify potential data quality problems and provide strong support for subsequent data cleaning and preprocessing. By optimizing the data quality of the Chinese and English bilingual corpus, the technical solution can indirectly improve the accuracy and fluency of machine translation. This is of great significance for improving the efficiency and quality of cross-language communication. The technical solution enhances the robustness of the system by introducing data quality evaluation and anomaly detection mechanisms. Even in the face of complex and changing data environments, the system can maintain stable performance and accurate output.

[0112] In summary, this technical solution provides strong support for Chinese-English bidirectional machine translation testing through refined data processing and comprehensive data quality assessment, which helps to improve the accuracy of machine translation and the overall robustness of the system.

[0113] Specifically, when the first abnormal evaluation coefficient does not exceed the preset first coefficient threshold, the semantic weight value of the sentence where the inconsistent, invalid and missing data is located is retrieved, and the data quality is secondary detected according to the semantic weight value of the sentence where the inconsistent, invalid and missing data is located, including:

[0114] Extracting the semantic weight value of the sentence where each inconsistent data is located, wherein the inconsistent data includes Chinese and English sentences that do not conform to the language features;

[0115] Extract the semantic weight value of each sentence containing invalid data;

[0116] Extracting the semantic weight value of each sentence where the missing data is located; wherein the sentence where the missing data is located is a sentence where words or characters are missing;

[0117] The second abnormality evaluation coefficient is obtained by using the semantic weight value of the sentence where each inconsistent data is located, the semantic weight value of the sentence where each invalid data is located, and the semantic weight value of the sentence where each missing data is located; wherein the second abnormality evaluation coefficient is obtained by the following formula:

[0118]

[0119] in

[0120]

[0121] Among them, P 01 Represents the second abnormality evaluation coefficient; N 01 、N 02 and N 03 Respectively represent the number of statements with inconsistent, invalid and missing data; W 01i , W 02i and W 03i Respectively represent the semantic weight values ​​corresponding to the sentences containing the i-th inconsistent, invalid and missing data; W 01h , W 02h and W 03h Respectively represent the semantic weights and values ​​of the sentences containing inconsistent, invalid and missing data; k 01 , k 02 and k 03 represent the first adjustment factor, the second adjustment factor and the third adjustment factor respectively, and the value ranges of the first adjustment factor, the second adjustment factor and the third adjustment factor are: [0.3, 2], [0.6, 3] and [0.8, 1.7];

[0122] The technical effect of the above technical solution is: by extracting the semantic weight values ​​of the sentences containing inconsistent, invalid and missing data, the technical solution can detect the data quality more finely. It not only focuses on the quantity of data, but also deeply considers the semantic importance of the sentences in which the data is located, so as to more comprehensively evaluate the data quality. By performing secondary detection using the semantic weight value, the technical solution can more accurately identify potential data quality issues. Even if the quantity of certain data does not exceed the preset threshold, it may still be regarded as an important problem point if its semantic weight is high.

[0123] By introducing the first adjustment factor, the second adjustment factor and the third adjustment factor, the technical solution can be flexibly adjusted according to different application scenarios and data characteristics. This allows it to maintain high accuracy and effectiveness when facing different data quality and semantic environments. Through the secondary detection mechanism, the technical solution can identify potential data quality issues at an early stage, thereby avoiding wasting time and resources in subsequent data processing and analysis. This helps to improve the overall data processing efficiency and quality. By optimizing the data quality of the Chinese and English bilingual corpus, the technical solution can indirectly enhance the robustness of the machine translation system. It helps to reduce translation errors or inaccuracies caused by data quality issues, thereby improving the accuracy and reliability of machine translation.

[0124] In summary, this technical solution provides strong support for Chinese-English bidirectional machine translation testing through refined data quality detection, accurate potential problem identification, flexible adjustment factor adaptation, improved data processing efficiency, and enhanced machine translation system robustness.

[0125] S2: According to the test requirements of Chinese-English bidirectional machine translation, a Chinese-English bidirectional machine translation test model based on grammatical composition analysis is constructed, and the Chinese-English bidirectional machine translation test model is evaluated and optimized to determine the optimal Chinese-English bidirectional machine translation test model;

[0126] In this embodiment, a Chinese-English bidirectional machine translation test model based on grammatical composition analysis is constructed, and the following operations are performed:

[0127] Obtain a set of Chinese and English bilingual corpora based on grammatical composition analysis;

[0128] Divide the Chinese-English bilingual corpus association set based on grammatical composition analysis to determine the Chinese-English bidirectional machine translation training set and the Chinese-English bidirectional machine translation test set;

[0129] According to the requirements of Chinese-English bidirectional machine translation test, select a neural network framework suitable for Chinese-English bidirectional machine translation test;

[0130] Based on the Chinese-English bidirectional machine translation training set, train the selected neural network framework suitable for Chinese-English bidirectional machine translation testing;

[0131] A Chinese-English bidirectional machine translation test model based on grammatical composition analysis was determined.

[0132] In this embodiment, the Chinese-English bidirectional machine translation test model is evaluated and optimized by performing the following operations:

[0133] Obtain a Chinese-English bidirectional machine translation test model based on grammatical composition analysis;

[0134] Based on the Chinese-English bidirectional machine translation test set, the Chinese-English bidirectional machine translation test model based on grammatical composition analysis was evaluated and tested;

[0135] Determine the evaluation test results based on the Chinese-English bidirectional machine translation test model;

[0136] Conduct mining and analysis on the evaluation test results based on the Chinese-English bidirectional machine translation test model;

[0137] Determine the optimization adjustment plan based on the Chinese-English bidirectional machine translation test model;

[0138] Optimize and adjust the Chinese-English bidirectional machine translation test model according to the optimization adjustment plan based on the Chinese-English bidirectional machine translation test model;

[0139] Determine the optimal Chinese-English bidirectional machine translation test model.

[0140] S3: Translate the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model, and determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis;

[0141] In this embodiment, the machine translation demand data is translated based on the optimal Chinese-English bidirectional machine translation test model, and the following operations are performed:

[0142] According to the requirements of Chinese-English bidirectional machine translation testing, obtain machine translation demand data;

[0143] Input machine translation demand data into the optimal Chinese-English bidirectional machine translation test model;

[0144] Translate the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model;

[0145] Determine the Chinese-English bidirectional machine translation results based on grammatical composition analysis;

[0146] Compare and analyze the Chinese-English bidirectional machine translation results with the preset Chinese-English bidirectional translation standards;

[0147] Determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis.

[0148] In this embodiment, the Chinese-English bidirectional machine translation test results based on grammatical composition analysis include:

[0149] When the Chinese-English bidirectional machine translation result matches the preset Chinese-English bidirectional translation standard, the Chinese-English bidirectional machine translation test result based on grammatical composition analysis is that the Chinese-English bidirectional machine translation is normal;

[0150] When the Chinese-English bidirectional machine translation result does not match the preset Chinese-English bidirectional translation standard, the Chinese-English bidirectional machine translation test result based on grammatical composition analysis is that the Chinese-English bidirectional machine translation is abnormal.

[0151] It should be noted that the Chinese-English bidirectional machine translation results were compared and analyzed with the preset Chinese-English bidirectional translation standards. The Chinese-English bidirectional machine translation test results based on grammatical composition analysis are shown in Table 1:

[0152] Table 1: Test results of Chinese-English bidirectional machine translation based on grammatical composition analysis

[0153]

[0154] Therefore, by comparing and analyzing the Chinese-English bidirectional machine translation results with the preset Chinese-English bidirectional translation standards, we can determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis. Based on the Chinese-English bidirectional machine translation test results, we can formulate a Chinese-English bidirectional machine translation test optimization and control plan. Based on the Chinese-English bidirectional machine translation test optimization and control plan, the optimal Chinese-English bidirectional machine translation test model can be intelligently optimized and controlled to meet the Chinese-English bidirectional machine translation needs.

[0155] S4: Based on the Chinese-English bidirectional machine translation test optimization and control solution, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled to meet the needs of Chinese-English bidirectional machine translation.

[0156] In this embodiment, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and managed, and the following operations are performed:

[0157] Obtain Chinese-English bidirectional machine translation test results based on grammatical composition analysis;

[0158] Conduct in-depth mining and analysis of Chinese-English bidirectional machine translation test results based on grammatical composition analysis;

[0159] Determine the optimization and control plan for Chinese-English bidirectional machine translation testing based on grammatical composition analysis;

[0160] Based on the Chinese-English bidirectional machine translation test optimization and control solution, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled;

[0161] The model structure and parameters of the optimal Chinese-English bidirectional machine translation test model are continuously adjusted, and multiple iterations are performed to determine a Chinese-English bidirectional machine translation test model that can meet the needs of Chinese-English bidirectional machine translation.

[0162] Therefore, by collecting Chinese-English bilingual corpora, processing the Chinese-English bilingual corpora, determining the Chinese-English bilingual corpus association set based on grammatical composition analysis, constructing a Chinese-English bidirectional machine translation test model based on grammatical composition analysis according to the Chinese-English bidirectional machine translation test requirements, evaluating and optimizing the Chinese-English bidirectional machine translation test model, determining the optimal Chinese-English bidirectional machine translation test model, translating the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model, determining the Chinese-English bidirectional machine translation test results based on grammatical composition analysis, formulating a Chinese-English bidirectional machine translation test optimization management and control plan based on the Chinese-English bidirectional machine translation test results, and intelligently optimizing and controlling the optimal Chinese-English bidirectional machine translation test model based on the Chinese-English bidirectional machine translation test optimization management and control plan to meet the Chinese-English bidirectional machine translation requirements, and performing bidirectional translation between Chinese and English based on grammatical composition analysis, so that the accuracy of Chinese-English bidirectional machine translation is high, and the quality of Chinese-English bidirectional machine translation can be improved.

[0163] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0164] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A Chinese-English bidirectional machine translation testing method based on grammatical composition analysis, characterized in that: The steps include: S1: Collect Chinese and English bilingual corpora, process them, and determine the Chinese and English bilingual corpus association set based on grammatical composition analysis; S2: According to the test requirements of Chinese-English bidirectional machine translation, a Chinese-English bidirectional machine translation test model based on grammatical composition analysis is constructed, and the Chinese-English bidirectional machine translation test model is evaluated and optimized to determine the optimal Chinese-English bidirectional machine translation test model; S3: Translate the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model, and determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis; S4: Based on the Chinese-English bidirectional machine translation test optimization and control solution, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled to meet the needs of Chinese-English bidirectional machine translation.

2. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 1 is characterized in that: In S1, a Chinese-English bilingual corpus is collected, and the following operations are performed: Acquire Chinese materials in different fields and topics to determine Chinese corpus data; Obtain English materials in different fields and topics to determine English corpus data; Among them, Chinese-English bilingual corpus data are determined based on Chinese corpus data and English corpus data.

3. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 2 is characterized in that: In S1, the Chinese-English bilingual corpus is processed by performing the following operations: Obtain bilingual corpus data in Chinese and English; Based on the sequential retrieval method, the Chinese and English bilingual corpus data are retrieved; Remove inconsistent, invalid and missing data in the Chinese-English bilingual corpus data that are of no value for Chinese-English bidirectional machine translation testing; Identify bilingual corpus data that is valuable for Chinese-English bidirectional machine translation testing; Obtain Chinese-English bilingual corpus data that is valuable for Chinese-English bidirectional machine translation testing; Perform word segmentation, part-of-speech tagging, syntax and grammar analysis on the valuable Chinese-English bilingual corpus data for Chinese-English bidirectional machine translation tests to make the source language and target language semantically matched; Determine the associated set of Chinese and English bilingual corpora based on grammatical composition analysis.

4. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 3 is characterized in that: In S1, the Chinese-English bilingual corpus is processed, and further includes: Extracting the amount of inconsistent data information that is worthless for Chinese-English bidirectional machine translation testing from the English bilingual corpus data as the first data amount; Extracting the amount of invalid data information that is worthless to the Chinese-English bidirectional machine translation test from the English bilingual corpus data as the second data amount; Extracting the amount of data information of sentences containing missing data that is worthless for Chinese-English bidirectional machine translation testing from the English bilingual corpus data as the third amount of data; wherein the sentence containing the missing data is a sentence with missing words or characters; The first abnormality evaluation coefficient is obtained by using the first data amount, the second data amount and the third data amount, wherein the first abnormality evaluation coefficient is obtained by the following formula: Among them, P 01 Denotes the first abnormal evaluation coefficient; D 01 , D 02 and D 03 Respectively represent the first data volume, the second data volume and the third data volume; α, β and γ represent the first adjustment coefficient, the second adjustment coefficient and the third adjustment coefficient; and the value ranges of the first adjustment coefficient, the second adjustment coefficient and the third adjustment coefficient are: [0.1, 5], [0.3, 2] and [0.2, 1.3]; Comparing the first abnormality evaluation coefficient with a preset first coefficient threshold; When the first abnormal evaluation coefficient exceeds a preset first coefficient threshold, it is determined that the English bilingual corpus data has data quality abnormality, and a data quality abnormality alarm is issued; When the first abnormality evaluation coefficient does not exceed the preset first coefficient threshold, the semantic weight value of the sentence containing inconsistent, invalid and missing data is retrieved, and the data quality is secondary checked based on the semantic weight value of the sentence containing inconsistent, invalid and missing data.

5. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 4 is characterized in that: When the first abnormal evaluation coefficient does not exceed the preset first coefficient threshold, the semantic weight value of the sentence where the inconsistent, invalid and missing data is located is retrieved, and the data quality is secondary detected according to the semantic weight value of the sentence where the inconsistent, invalid and missing data is located, including: Extracting the semantic weight value of the sentence where each inconsistent data is located, wherein the inconsistent data includes Chinese and English sentences that do not conform to the language features; Extract the semantic weight value of each sentence containing invalid data; Extracting the semantic weight value of each sentence where the missing data is located; wherein the sentence where the missing data is located is a sentence where words or characters are missing; The second abnormality evaluation coefficient is obtained by using the semantic weight value of the sentence where each inconsistent data is located, the semantic weight value of the sentence where each invalid data is located, and the semantic weight value of the sentence where each missing data is located; wherein the second abnormality evaluation coefficient is obtained by the following formula: in Among them, P 01 Represents the second abnormality evaluation coefficient; N 01 、N 02 and N 03 Respectively represent the number of statements with inconsistent, invalid and missing data; W 01i , W 02i and W 03i Respectively represent the semantic weight values ​​corresponding to the sentences containing the i-th inconsistent, invalid and missing data; W 01h , W 02h and W 03h Respectively represent the semantic weights and values ​​of the sentences containing inconsistent, invalid and missing data; k 01 , k 02 and k 03 represent the first adjustment factor, the second adjustment factor and the third adjustment factor respectively, and the value ranges of the first adjustment factor, the second adjustment factor and the third adjustment factor are: [0.3, 2], [0.6, 3] and [0.8, 1.7]; comparing the second abnormality evaluation coefficient with a preset second coefficient threshold; When the second abnormality evaluation coefficient exceeds a preset second coefficient threshold, it is determined that the English bilingual corpus data has data quality abnormality, and a data quality abnormality alarm is issued.

6. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 3 is characterized in that: In S2, a Chinese-English bidirectional machine translation test model based on grammatical composition analysis is constructed, and the following operations are performed: Obtain a set of Chinese and English bilingual corpora based on grammatical composition analysis; Divide the Chinese-English bilingual corpus association set based on grammatical composition analysis to determine the Chinese-English bidirectional machine translation training set and the Chinese-English bidirectional machine translation test set; According to the requirements of Chinese-English bidirectional machine translation test, select a neural network framework suitable for Chinese-English bidirectional machine translation test; Based on the Chinese-English bidirectional machine translation training set, train the selected neural network framework suitable for Chinese-English bidirectional machine translation testing; A Chinese-English bidirectional machine translation test model based on grammatical composition analysis was determined.

7. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 6 is characterized in that: In S2, the Chinese-English bidirectional machine translation test model is evaluated and optimized by performing the following operations: Obtain a Chinese-English bidirectional machine translation test model based on grammatical composition analysis; Based on the Chinese-English bidirectional machine translation test set, the Chinese-English bidirectional machine translation test model based on grammatical composition analysis was evaluated and tested; Determine the evaluation test results based on the Chinese-English bidirectional machine translation test model; Conduct mining and analysis on the evaluation test results based on the Chinese-English bidirectional machine translation test model; Determine the optimization adjustment plan based on the Chinese-English bidirectional machine translation test model; Optimize and adjust the Chinese-English bidirectional machine translation test model according to the optimization adjustment plan based on the Chinese-English bidirectional machine translation test model; Determine the optimal Chinese-English bidirectional machine translation test model.

8. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 7 is characterized in that: In S3, the machine translation demand data is translated based on the optimal Chinese-English bidirectional machine translation test model, and the following operations are performed: According to the requirements of Chinese-English bidirectional machine translation testing, obtain machine translation demand data; Input machine translation demand data into the optimal Chinese-English bidirectional machine translation test model; Translate the machine translation demand data based on the optimal Chinese-English bidirectional machine translation test model; Determine the Chinese-English bidirectional machine translation results based on grammatical composition analysis; Compare and analyze the Chinese-English bidirectional machine translation results with the preset Chinese-English bidirectional translation standards; Determine the Chinese-English bidirectional machine translation test results based on grammatical composition analysis.

9. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 8, characterized in that: In S3, the Chinese-English bidirectional machine translation test results based on grammatical composition analysis include: When the Chinese-English bidirectional machine translation result matches the preset Chinese-English bidirectional translation standard, the Chinese-English bidirectional machine translation test result based on grammatical composition analysis is that the Chinese-English bidirectional machine translation is normal; When the Chinese-English bidirectional machine translation result does not match the preset Chinese-English bidirectional translation standard, the Chinese-English bidirectional machine translation test result based on grammatical composition analysis is that the Chinese-English bidirectional machine translation is abnormal.

10. The Chinese-English bidirectional machine translation testing method based on grammatical composition analysis according to claim 9, characterized in that: In S4, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled, and the following operations are performed: Obtain Chinese-English bidirectional machine translation test results based on grammatical composition analysis; Conduct in-depth mining and analysis of Chinese-English bidirectional machine translation test results based on grammatical composition analysis; Determine the optimization and control plan for Chinese-English bidirectional machine translation testing based on grammatical composition analysis; Based on the Chinese-English bidirectional machine translation test optimization and control solution, the optimal Chinese-English bidirectional machine translation test model is intelligently optimized and controlled; The model structure and parameters of the optimal Chinese-English bidirectional machine translation test model are continuously adjusted, and multiple iterations are performed to determine a Chinese-English bidirectional machine translation test model that can meet the needs of Chinese-English bidirectional machine translation.

Images