English translation text data processing method and system and storage medium

Abstract

The invention relates to the technical field of text data processing, in particular to an English translation text data processing method and system and a storage medium, the system comprises a data processing center, a text translation unit, a translation quality unit, a paragraph management unit, an error locking unit and a back-end feedback unit; through term consistency check, the uniformity of term translation in the professional field is ensured, the professionality and accuracy of translation are improved, and meanwhile, translation quality analysis is deeply performed on the translation text, so that whether the overall translation quality of the target text data is qualified or not can be visually known according to the information feedback condition; and meanwhile, paragraphs with unqualified translation quality are analyzed in an information progressive manner, so that targeted translation management is performed on the paragraphs, and parameters in the translation tool are adjusted based on a character feedback condition, namely, reasonable adjustment is performed aiming at defects, so that the translation quality of the translation tool on target text data is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of text data processing, and in particular to a method, system and storage medium for processing English translation text data. Background Art

[0002] With the deepening of international communication, the demand for English document translation is also increasing, prompting the emergence of a large number of English translation tools. These English translation tools are generally divided into online and local versions. Both online and local versions search for the translation in the database for translation. The emergence of these translation tools has greatly satisfied the translation needs of users, contributed to improving translation efficiency and promoting social progress.

[0003] There are several problems with processing English translation text data: Traditional translation text data processing methods struggle to fully match the sentences to be translated. They typically perform a one-to-one word-by-word translation of the sentences to be translated, often leading to errors in tense and word order. The translation is also awkward, and translation quality cannot be guaranteed. Furthermore, it is difficult to manage paragraphs in the translated text in a targeted manner, resulting in poor readability and fluency, hindering its use. Furthermore, it is difficult to make targeted adjustments to errors in the translated text, reducing the efficiency and quality of the translation tools.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide an English translation text data processing method, system and storage medium to solve the technical defects mentioned above. The present invention ensures the uniformity of professional field term translation through term consistency check, reduces the situation where multiple translation versions of the same term appear in the document, improves the professionalism and accuracy of the translation, and at the same time performs in-depth translation quality analysis on the translated text, so as to intuitively understand whether the overall translation quality of the target text data is qualified based on the information feedback, so as to rationally classify and manage the target text data to improve the management efficiency of the target text data after translation. At the same time, paragraphs with low translation quality are analyzed in an information progressive manner, so as to intuitively understand the translation quality of each paragraph in the target text data, so as to perform targeted translation management on the paragraphs, and at the same time, the parameters in the translation tool are adjusted based on the text feedback, that is, rational adjustments are made to the defects to improve the translation quality of the target text data by the translation tool.

[0006] The purpose of the present invention can be achieved by the following technical solution: A method for processing English translation text data, comprising the following steps:

[0007] S1: Obtain the text uploaded by the user, and obtain the corresponding translated initial translation text based on the text uploaded by the user;

[0008] S2: Preprocess the initial translation text;

[0009] S3: Perform terminology consistency check on the pre-processed initial translation text based on the preset terminology database, and replace inconsistent terms with corresponding standard terms in the preset terminology database;

[0010] S4: Perform a multi-dimensional evaluation and analysis of the data processing quality of the replaced translation text, and perform a discriminative processing on the obtained multi-dimensional translation quality score. If a qualified signal is obtained, the replaced translation text is output as feedback. If an unqualified signal is obtained, proceed to S5 and S6;

[0011] S5: Based on the segment translation quality classification and marking analysis under the unqualified signal, interactive analysis is performed on the obtained segment defect data, segment normal data, low-scoring segments, and high-scoring segments, and the obtained retranslated segments, adjusted segments, and qualified segments are output as feedback;

[0012] S6: Based on the error locking and targeted regulation analysis under the unqualified signal, the obtained error frequency value PFm is discriminated and processed, and the obtained translation defect type is output as feedback.

[0013] Preferably, the terminology consistency check process is as follows: extract terms from the preprocessed initial translation text, match the extracted terms with the standard terms in the terminology database, and if the matches are inconsistent, replace the term with the corresponding standard term in the preset terminology database. If the matches are consistent, the term is not replaced, and the replaced translation text is obtained.

[0014] Preferably, the data processing quality multi-dimensional evaluation and analysis process is as follows:

[0015] The replaced translated text is set as the target text data, a preset translation quality scoring model is retrieved from the server, the target text data is input into the translation quality scoring model, and the output results of the translation quality scoring model are obtained. The output results represent an accuracy score, a fluency score, and a readability score. Then, an average value of the output results of the translation quality scoring model is obtained, and the average value of the output results is set as the automatic translation quality score;

[0016] At the same time, n professional translators are randomly selected to perform manual translation quality ratings on the target text data, where n is a natural number greater than 3. A set A of manual translation quality ratings is constructed, and the maximum and minimum elements in set A are removed and then the average is calculated. The combination of the calculated averages is set as the average of the manual translation ratings;

[0017] A preset weight factor is assigned to the average of the automatic translation quality score and the manual translation score, and the sum of the automatic translation quality score and the average of the manual translation score multiplied by the corresponding preset weight factor is set as the multidimensional translation quality score. The multidimensional translation quality score is discriminated and processed to obtain an unqualified signal or a qualified signal.

[0018] Preferably, the segment translation quality classification and marking analysis process is as follows: based on the translation quality scoring model, an automatic translation quality score of the segment text data is obtained, and the automatic translation quality score is discriminated. If the automatic translation quality score is less than a preset automatic translation quality score threshold, the corresponding segment text data is set as segment defect data; if the automatic translation quality score is greater than or equal to the preset automatic translation quality score threshold, the corresponding segment text data is set as segment normal data.

[0019] Preferably, based on the manual translation quality scoring of the segment text data by professional translators, the average manual translation score of each segment text data is obtained, and the average manual translation score is subjected to discrimination processing. If the average manual translation score is less than a preset manual translation score average threshold, the corresponding segment text data is set as a low score segment; if the average manual translation score is greater than or equal to the preset manual translation score average threshold, the corresponding segment text data is set as a high score segment;

[0020] Interactive analysis is performed on segment defect data, segment normal data, low-scoring segments, and high-scoring segments to obtain retranslated paragraphs, adjusted paragraphs, and qualified paragraphs.

[0021] Preferably, the error locking and targeted regulation analysis process is as follows:

[0022] The retranslated paragraphs and the adjusted paragraphs are collectively referred to as abnormal paragraphs. The translation error type of each abnormal paragraph is obtained, and the parameter in the translation error type is set as the error parameter CWg, where g is a natural number greater than zero.

[0023] Obtain the ratio of the number of abnormal paragraphs corresponding to the error parameter CWg to the total number of abnormal paragraphs, and set it as the type ratio value LXg;

[0024] The type proportion value LXg is judged and processed. If the type proportion value LXg is greater than or equal to the preset type proportion value threshold, the corresponding error parameter is judged to be a frequent parameter. If the type proportion value LXg is less than the preset type proportion value threshold, the corresponding error parameter is judged to be a regular parameter.

[0025] Preferably, the total number of occurrences of each conventional parameter in the abnormal paragraph is obtained, and the total number of occurrences of each conventional parameter in the abnormal paragraph is set as the error frequency value PFm, where m is a natural number greater than zero. The error frequency value PFm is subjected to discrimination processing. If the error frequency value PFm is greater than or equal to a preset error frequency value threshold, the corresponding conventional parameter is judged to be a defective parameter. If the error frequency value PFm is less than the preset error frequency value threshold, the corresponding conventional parameter is judged to be an occasional parameter. Frequent parameters and defective parameters are collectively referred to as translation defect types.

[0026] The beneficial effects of the present invention are as follows:

[0027] (1) The present invention ensures the uniformity of professional terminology translation through terminology consistency checking, reduces the occurrence of multiple translation versions of the same term in a document, improves the professionalism and accuracy of translation, and at the same time conducts in-depth translation quality analysis on the translated text, so as to intuitively understand whether the overall translation quality of the target text data is qualified based on information feedback, so as to rationally classify and manage the target text data, thereby improving the management efficiency of the target text data after translation;

[0028] (2) The present invention analyzes paragraphs with low translation quality in an information progressive manner so as to intuitively understand the translation quality of each paragraph in the target text data, so as to carry out targeted translation management of the paragraphs. At the same time, the parameters in the translation tool are adjusted based on the text feedback, that is, reasonable adjustments are made to the defects to improve the translation quality of the target text data by the translation tool. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention will be further described below with reference to the accompanying drawings;

[0030] Figure 1 It is a reference schematic diagram of the method of the present invention;

[0031] Figure 2 It is a flow chart of the system of the present invention. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0033] References to "embodiments" herein mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments;

[0034] Example 1:

[0035] See also Figures 1 to 2 As shown, the present invention is a method for processing English translation text data, comprising the following steps:

[0036] S1: Obtain the text uploaded by the user, and obtain the corresponding translated initial translation text based on the text uploaded by the user;

[0037] S2: Preprocess the initial translation text, including word segmentation and part-of-speech tagging;

[0038] S3: Perform terminology consistency check on the pre-processed initial translation text based on the preset terminology database, and replace inconsistent terms with corresponding standard terms in the preset terminology database;

[0039] The terminology consistency check process is as follows: extract terms from the preprocessed initial translation text, match the extracted terms with the standard terms in the terminology database, and if the matches are inconsistent, replace the terms with the corresponding standard terms in the preset terminology database. If the matches are consistent, the terms are not replaced, and the replaced translation text is obtained.

[0040] S4: Perform a multi-dimensional evaluation and analysis of the data processing quality of the replaced translation text, and perform a discriminative processing on the obtained multi-dimensional translation quality score. If a qualified signal is obtained, the replaced translation text is output as feedback. If an unqualified signal is obtained, proceed to S5 and S6;

[0041] S5: Based on the segment translation quality classification and marking analysis under the unqualified signal, interactive analysis is performed on the obtained segment defect data, segment normal data, low-scoring segments, and high-scoring segments, and the obtained retranslated segments, adjusted segments, and qualified segments are output as feedback;

[0042] S6: Based on the error locking and targeted regulation analysis under the unqualified signal, the obtained error frequency value PFm is discriminated and processed, and the obtained translation defect type is output as feedback.

[0043] Example 2:

[0044] An English translation text data processing system includes a data processing center, a text translation unit, a translation quality unit, a paragraph management unit, an error locking unit, and a back-end feedback unit. The data processing center is connected to the text translation unit and the translation quality unit in a one-way communication manner. The translation quality unit is connected to the paragraph management unit in a one-way communication manner. The paragraph management unit is connected to the error locking unit in a one-way communication manner. The error locking unit is connected to the back-end feedback unit in a one-way communication manner.

[0045] The data processing center is used to retrieve the text uploaded by the user and send the text uploaded by the user to the text translation unit for text data translation processing supervision analysis to obtain a corrected translation text;

[0046] The translation quality unit is used to conduct multi-dimensional evaluation and analysis of the data processing quality of the revised translation text, so as to intuitively understand whether the overall translation quality of the target text data is qualified based on the information feedback, so as to rationally classify and manage the target text data and improve the management efficiency of the target text data after translation. The specific multi-dimensional evaluation and analysis process of data processing quality is as follows:

[0047] The replaced translated text is set as the target text data, a pre-set translation quality scoring model is retrieved from the server, the target text data is input into the translation quality scoring model, and the output results of the translation quality scoring model are obtained. The output results represent the accuracy score, fluency score, readability score, etc., and then the average value of the output results of the translation quality scoring model is obtained, and the average value of the output results is set as the automatic translation quality score;

[0048] At the same time, n professional translators are randomly selected to perform manual translation quality ratings (0-100) on the target text data, where n is a natural number greater than 3. A set A of manual translation quality ratings is constructed. The maximum and minimum elements in set A are removed and the average is calculated. The combination of the calculated averages is set as the average of the manual translation ratings;

[0049] A preset weight factor is assigned to the average of the automatic translation quality score and the manual translation score, and the sum of the automatic translation quality score and the average of the manual translation score multiplied by the corresponding preset weight factor is set as the multidimensional translation quality score. The multidimensional translation quality score is then discriminated:

[0050] If the translation quality multi-dimensional score is less than the preset translation quality multi-dimensional score threshold, a failure signal is generated;

[0051] If the translation quality multi-dimensional score is greater than or equal to the preset translation quality multi-dimensional score threshold, a qualified signal is generated;

[0052] The back-end feedback unit is used to respond to unqualified signals or qualified signals and immediately display the preset warning text corresponding to the unqualified signals or qualified signals, so as to intuitively understand whether the overall translation quality of the target text data is qualified based on the information feedback, so as to rationally classify and manage the target text data, thereby improving the management efficiency of the target text data after translation;

[0053] When an unqualified signal is generated, the paragraph management unit is used to perform segment translation quality classification and marking analysis on the segmented text data, so as to intuitively understand the translation quality of each paragraph in the target text data and perform targeted translation management on the paragraphs. The specific segment translation quality classification and marking analysis process is as follows:

[0054] Obtaining an automatic translation quality score for the segment text data based on the translation quality score model, and performing discriminative processing on the automatic translation quality score; if the automatic translation quality score is less than a preset automatic translation quality score threshold, setting the corresponding segment text data as segment defective data; if the automatic translation quality score is greater than or equal to the preset automatic translation quality score threshold, setting the corresponding segment text data as segment normal data;

[0055] At the same time, based on the manual translation quality scoring of the segment text data by professional translators, the average manual translation score of each segment text data is obtained, and the average manual translation score is subjected to discrimination processing. If the average manual translation score is less than the preset manual translation score average threshold, the corresponding segment text data is set to a low score segment; if the average manual translation score is greater than or equal to the preset manual translation score average threshold, the corresponding segment text data is set to a high score segment;

[0056] Interactive analysis of segment defect data, segment normal data, low segmentation and high segmentation:

[0057] If segment defect data and low-scoring segments are obtained, the corresponding segment text data is set as a retranslated segment;

[0058] If segment defect data and high-score segment or segment normal data and low-score segment are obtained, the corresponding segment text data is set as the adjustment paragraph;

[0059] If normal segment data and high-scoring segments are obtained, the corresponding segment text data is set as a qualified paragraph;

[0060] The back-end feedback unit is used to respond to retranslated paragraphs, adjusted paragraphs and qualified paragraphs, and immediately mark the retranslated paragraphs in red, the adjusted paragraphs in yellow, and the qualified paragraphs in green, so as to intuitively understand the translation quality of each paragraph in the target text data, so as to carry out targeted translation management of the paragraphs, such as retranslating the retranslated paragraphs and adjusting the sentences and fluency of the adjusted paragraphs, which helps to reduce the translation task volume of the target text data and improve the targeted processing efficiency of the target text data.

[0061] Example 3:

[0062] When a failure signal is generated, the error locking unit is used to perform error locking and targeted regulation analysis on the collected translation error types, so as to adjust the parameters in the translation tool based on the text feedback. In other words, rational adjustments are made to the defects to improve the translation quality of the target text data. The specific error locking and targeted regulation analysis process is as follows:

[0063] The retranslated paragraphs and adjusted paragraphs are collectively referred to as abnormal paragraphs. The translation error type in each abnormal paragraph is obtained, including semantic errors, grammatical errors, etc.

[0064] The parameter in the translation error type is set to the error parameter CWg, where g is a natural number greater than zero. For example, when g=1, the error parameter CW1 indicates a semantic error; when g=2, the error parameter CW2 indicates a syntactic error, and so on;

[0065] Obtain the ratio of the number of abnormal paragraphs corresponding to the error parameter CWg to the total number of abnormal paragraphs, and set it as the type ratio value LXg;

[0066] The type ratio value LXg is judged. If the type ratio value LXg is greater than or equal to the preset type ratio value threshold, the corresponding error parameter is judged to be a frequent parameter. If the type ratio value LXg is less than the preset type ratio value threshold, the corresponding error parameter is judged to be a regular parameter.

[0067] At the same time, the total number of occurrences of each regular parameter in the abnormal paragraph is obtained, and the total number of occurrences of each regular parameter in the abnormal paragraph is set as the error frequency value PFm, where m is a natural number greater than zero. For example, when m=1, the error frequency value PF1 represents the error frequency value of the first regular parameter; when m=2, the error frequency value PF2 represents the error frequency value of the second regular parameter, and so on;

[0068] The error frequency value PFm is also judged and processed. If the error frequency value PFm is greater than or equal to a preset error frequency value threshold, the corresponding regular parameter is judged to be a defective parameter. If the error frequency value PFm is less than the preset error frequency value threshold, the corresponding regular parameter is judged to be an occasional parameter. The frequent parameters and defect parameters are collectively referred to as translation defect types. The back-end feedback unit is used to respond to the translation defect type and immediately display the preset warning text corresponding to the translation defect type, so as to adjust the parameters in the translation tool based on the text feedback, that is, to make reasonable adjustments based on the defects, so as to improve the translation quality of the target text data by the translation tool;

[0069] A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements a method for processing English translation text data;

[0070] In summary, the present invention ensures the uniformity of professional terminology translation through terminology consistency check, reduces the situation where multiple translation versions of the same term appear in documents, improves the professionalism and accuracy of translation, and at the same time performs in-depth translation quality analysis on the translated text, so as to intuitively understand whether the overall translation quality of the target text data is qualified based on the information feedback, so as to rationally classify and manage the target text data, so as to improve the management efficiency of the target text data after translation, and at the same time analyzes paragraphs with low translation quality in an information progressive manner, so as to intuitively understand the translation quality of each paragraph in the target text data, so as to perform targeted translation management of the paragraphs, and at the same time adjust the parameters in the translation tool based on the text feedback, that is, make rational adjustments to the defects, so as to improve the translation quality of the target text data by the translation tool.

[0071] The threshold is set for result comparison and analysis to determine whether it is good or bad. The value of the threshold is set based on a combination of large-scale model analysis of sample data and manual experience to enter and store data. It can also be appropriately adjusted based on seasonal or common sense influencing conditions.

[0072] The size of the coefficient is to quantify each parameter to obtain a specific numerical value, which is convenient for subsequent comparison. The size of the coefficient depends on the amount of sample data and the preliminary setting of the corresponding operating coefficient for each set of sample data by technical personnel in this field; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0073] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for processing English translation text data, characterized in that: The following steps are involved: S1: Obtain the text uploaded by the user, and obtain the corresponding translated initial translation text based on the text uploaded by the user; S2: Preprocess the initial translation text; S3: Perform terminology consistency check on the pre-processed initial translation text based on the preset terminology database, and replace inconsistent terms with corresponding standard terms in the preset terminology database; S4: Perform a multi-dimensional evaluation and analysis of the data processing quality of the replaced translation text, and perform a discriminative processing on the obtained multi-dimensional translation quality score. If a qualified signal is obtained, the replaced translation text is output as feedback. If an unqualified signal is obtained, proceed to S5 and S6; S5: Based on the segment translation quality classification and marking analysis under the unqualified signal, interactive analysis is performed on the obtained segment defect data, segment normal data, low-scoring segments, and high-scoring segments, and the obtained retranslated segments, adjusted segments, and qualified segments are output as feedback; S6: Based on the error locking and targeted regulation analysis under the unqualified signal, the obtained error frequency value PFm is discriminated and processed, and the obtained translation defect type is output as feedback.

2. The method for processing English translation text data according to claim 1, characterized in that ,The term consistency checking process is as follows: extract terms from the preprocessed initial ,translation text, and match the extracted terms with the standard ,terms in the term library. If the matches are inconsistent, the ,terms that need to be replaced are replaced with the corresponding ,standard terms in the preset term library. If the matches are consistent, ,they are not replaced and the replaced translation text is ,obtained.

3. A method for processing English translation text data, characterized in that: The multi-dimensional evaluation and analysis process of data processing quality is as follows: The replaced translated text is set as the target text data, a preset translation quality scoring model is retrieved from the server, the target text data is input into the translation quality scoring model, and the output results of the translation quality scoring model are obtained. The output results represent an accuracy score, a fluency score, and a readability score. Then, an average value of the output results of the translation quality scoring model is obtained, and the average value of the output results is set as the automatic translation quality score; At the same time, n professional translators are randomly selected to perform manual translation quality ratings on the target text data, where n is a natural number greater than 3. A set A of manual translation quality ratings is constructed, and the maximum and minimum elements in set A are removed and then the average is calculated. The combination of the calculated averages is set as the average of the manual translation ratings; A preset weight factor is assigned to the average of the automatic translation quality score and the manual translation score, and the sum of the automatic translation quality score and the average of the manual translation score multiplied by the corresponding preset weight factor is set as the multidimensional translation quality score. The multidimensional translation quality score is discriminated and processed to obtain an unqualified signal or a qualified signal.

4. The method for processing English translation text data according to claim 1, wherein: The segment translation quality classification and marking analysis process is as follows: based on the translation quality scoring model, an automatic translation quality score of the segment text data is obtained, and the automatic translation quality score is discriminated. If the automatic translation quality score is less than a preset automatic translation quality score threshold, the corresponding segment text data is set as segment defect data; if the automatic translation quality score is greater than or equal to the preset automatic translation quality score threshold, the corresponding segment text data is set as segment normal data.

5. The method for processing English translation text data according to claim 4, wherein: Based on the manual translation quality scoring of the segment text data by professional translators, the average manual translation score of each segment text data is obtained, and the average manual translation score is discriminated. If the average manual translation score is less than the preset manual translation score average threshold, the corresponding segment text data is set to a low score segment; if the average manual translation score is greater than or equal to the preset manual translation score average threshold, the corresponding segment text data is set to a high score segment; Interactive analysis is performed on segment defect data, segment normal data, low-scoring segments, and high-scoring segments to obtain retranslated paragraphs, adjusted paragraphs, and qualified paragraphs.

6. The method for processing English translation text data according to claim 1, wherein: The error locking and targeted regulation analysis process is as follows: The retranslated paragraphs and the adjusted paragraphs are collectively referred to as abnormal paragraphs. The translation error type of each abnormal paragraph is obtained, and the parameter in the translation error type is set as the error parameter CWg, where g is a natural number greater than zero. Obtain the ratio of the number of abnormal paragraphs corresponding to the error parameter CWg to the total number of abnormal paragraphs, and set it as the type ratio value LXg; The type proportion value LXg is judged and processed. If the type proportion value LXg is greater than or equal to the preset type proportion value threshold, the corresponding error parameter is judged to be a frequent parameter. If the type proportion value LXg is less than the preset type proportion value threshold, the corresponding error parameter is judged to be a regular parameter.

7. The method for processing English translation text data according to claim 6, characterized in that: The total number of occurrences of each regular parameter in the abnormal paragraph is obtained, and the total number of occurrences of each regular parameter in the abnormal paragraph is set as the error frequency value PFm, where m is a natural number greater than zero. The error frequency value PFm is judged. If the error frequency value PFm is greater than or equal to the preset error frequency value threshold, the corresponding regular parameter is judged to be a defective parameter. If the error frequency value PFm is less than the preset error frequency value threshold, the corresponding regular parameter is judged to be an occasional parameter. Frequent parameters and defective parameters are collectively referred to as translation defect types.

8. An English translation text data processing system, the system being applied to an English translation text data processing method according to any one of claims 1 to 7, characterized in that: It includes data processing center, text translation unit, translation quality unit, paragraph management unit, error locking unit and back-end feedback unit; The data processing center is used to retrieve the text uploaded by the user and send the text uploaded by the user to the text translation unit for text data translation processing supervision analysis to obtain a corrected translation text; The translation quality unit is used to conduct multi-dimensional evaluation and analysis of the data processing quality of the revised translation text to obtain unqualified or qualified signals; When an unqualified signal is generated, the paragraph management unit is used to perform segment translation quality classification and marking analysis on the divided segment text data to obtain re-translated paragraphs, adjusted paragraphs and qualified paragraphs. The error locking unit is used to perform error locking and targeted regulation analysis on the collected translation error types to obtain translation defect types.

9. A computer-readable storage medium, characterized in that An English translation text data processing method according to any one of claims 1 to 7 is stored and can be loaded and executed by a processor.

Images