English automatic evaluation system based on speech recognition

Abstract

The invention discloses an English automatic evaluation system based on speech recognition, and relates to the technical field of speech recognition and evaluation, and the system comprises a data collection end, a speech recognition end and a memory practice end, and the data collection end and the speech recognition end are jointly provided with an abnormity feedback module. The data collection end is used for obtaining and updating data related to English evaluation in real time, carrying out grading and classification processing on the data according to the difficulty level and the type of the data, and monitoring the data at the same time. According to the automatic English evaluation system based on speech recognition, during English evaluation, the phenomena of evaluation data errors and repetition are avoided, the accuracy of evaluation results is improved, a real language communication environment is simulated through virtual AI cartoon characters, a face-to-face communication scene is achieved, the practical oral application ability of examinees can be better checked, and the evaluation efficiency is improved. The English ability of student users can be improved through the system.

Description

Technical Field

[0001] The present invention relates to the technical field of speech recognition and evaluation, and in particular to an English automatic evaluation system based on speech recognition. Background Art

[0002] Speech assessment technology, also known as spoken language assessment or automated spoken language evaluation, is generally understood as computer-assisted language learning, where a machine automatically scores pronunciation. This technology can overcome the time-consuming, labor-intensive, and inefficient nature of manual assessment.

[0003] In English assessments, the existing English assessment materials are obtained from the Internet and books. However, the test information on the Internet and books is not necessarily completely correct and may contain errors. In addition, duplication of information is also prone to occur when obtaining it, resulting in inaccurate test questions in the assessment system and thus inaccurate assessment results.

[0004] Existing English assessment methods rely on students recording their voices into the assessment system through a microphone. This simply captures the voice, resulting in an unrealistic test scenario and a failure to demonstrate students' actual oral adaptability, which in turn affects test results.

[0005] After the English assessment is completed, the existing assessment system will give a scoring result. After the scoring is completed, student users cannot directly practice the wrong questions or extend the wrong questions on the assessment system to improve their English ability.

[0006] Therefore, an automatic English assessment system based on speech recognition is proposed to solve the above problems. Summary of the Invention

[0007] The main purpose of the present invention is to provide an English automatic evaluation system based on speech recognition to solve the problems raised in the above background.

[0008] To achieve the above-mentioned object, the technical solution adopted by the present invention is: an automatic English assessment system based on speech recognition, comprising a data collection end, a speech recognition end and a memory practice end, wherein the data collection end and the speech recognition end are jointly provided with an abnormal feedback module;

[0009] The data collection terminal is used to obtain and update data related to English assessment in real time, and classify and process the data according to the difficulty and type of the data. At the same time, the data is monitored to determine whether there are errors and duplications in the data. If errors and duplications are found, they can be corrected automatically based on the information on the Internet. If they cannot be corrected automatically, an exception is generated and reported to the system for feedback;

[0010] The voice recognition terminal is used to receive the voice signal of the student user to obtain a voice evaluation package, and at the same time obtain a video evaluation package through face-to-face communication between the virtual AI cartoon character and the student user, recognize the voice signals in the voice evaluation package and the video evaluation package, and perform amplification, filtering and noise reduction processing at the same time, automatically compare and judge the voice signal with the data, and automatically give the judgment result and corresponding score. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still an objection, the student user can enter the complaint exception and report the system for feedback;

[0011] The memory practice end is used to receive the judgment results through comparison and form an electronic wrong question collection. At the same time, it analyzes the types of wrong questions according to the electronic wrong question collection, obtains a large amount of materials of the same level and type from the materials according to the types of wrong questions, and stores the materials. Student users can practice multiple times from the materials to master the difficult points of English.

[0012] The abnormality feedback module provides feedback when an abnormality occurs in the system, and feeds back the abnormality information to technical personnel for system improvement.

[0013] The data collection terminal includes a data acquisition module, a data monitoring module and a data correction module;

[0014] The data acquisition module includes an English assessment data unit, a data update unit and a grading and classification unit;

[0015] The English assessment data unit is used to store all relevant data for English assessment such as test papers, websites and books through a storage device;

[0016] The data updating unit is used to update data in real time via the Internet;

[0017] The grading and classification unit is used to grade the data according to the degree of difficulty and classify the data according to the type of data.

[0018] The data monitoring module includes a data comparison unit, a data duplication unit and a data judgment unit;

[0019] The data comparison unit is used to compare the acquired and real-time updated data with English grammar, vocabulary and sentences and obtain a first comparison result, which is divided into two categories according to errors and repetitions;

[0020] The data duplication unit is used to screen each evaluation test question in the acquired and real-time updated data, and record the repeated test questions as repeated test questions;

[0021] The data judgment unit is used to reconfirm the first comparison result, upload the correct data to the data collection terminal, and record the erroneous and repeated data as the second comparison result.

[0022] The data correction module includes a data search unit and a correction execution unit;

[0023] The data search unit is used to collect relevant content on the Internet based on the content of the second comparison result. If the second comparison result is a word, the pronunciation, letter composition and Chinese translation of the word are collected on the Internet. If the second comparison result is a sentence, the word composition, letter composition, Chinese translation and grammar of the sentence are collected on the Internet.

[0024] The execution correction unit is used to correct the repeated test questions and the second comparison result, delete the repeated test questions, and modify the second comparison result according to Internet information.

[0025] The speech recognition terminal includes a speech receiving module, a virtual evaluation module, a speech processing module and a judgment and scoring module;

[0026] The voice receiving module includes a voice receiving unit and a voice storage unit;

[0027] The voice receiving unit is used to receive the voice of the student user through a microphone;

[0028] The voice storage unit is used to save the voice of the student user to obtain a voice evaluation package.

[0029] The virtual evaluation module includes a virtual communication unit and a video storage unit;

[0030] The virtual communication unit is used for the virtual AI cartoon character to conduct evaluation and communication with the student user through the camera and generate a video;

[0031] The video storage unit is used to save the video to obtain a video evaluation package.

[0032] The speech processing module includes a speech signal extraction unit, a speech amplification unit and a speech filtering and noise reduction unit;

[0033] The voice signal extraction unit is used to identify the voice signals in the voice evaluation package and the video evaluation package and obtain the voice signals;

[0034] The voice amplification unit is used to amplify the sound of the voice signal;

[0035] The speech filtering and noise reduction unit is used to remove or weaken inaccurate signal frequencies and interference components in the speech signal, thereby obtaining cleaner and more accurate information.

[0036] The judgment and scoring module includes an evaluation and comparison unit, an evaluation and scoring unit, and an evaluation objection unit;

[0037] The comparison and evaluation unit compares the voice signal with the text and voice information in the data and displays the comparison result;

[0038] The evaluation and scoring unit is used to score the speech signal from 0 to 100 points based on the comparison result, obtain the scoring result and display the scoring information;

[0039] The evaluation objection unit is used for students to file an appeal when they have objections to the scoring results.

[0040] The memory practice terminal includes a wrong question acquisition module and an English practice module;

[0041] The wrong question acquisition module includes a wrong question receiving unit, a similar question acquisition unit and a wrong question collection unit;

[0042] The wrong question receiving unit is used to receive the wrong questions in the comparison result and perform a comprehensive analysis on the wrong questions;

[0043] The same type acquisition unit is used to search and acquire the same type of evaluation materials for wrong questions;

[0044] The wrong question collection unit is used to collect and store the wrong questions after each evaluation as a first wrong question collection, and to collect and store the obtained evaluation materials of the same type as a second wrong question collection.

[0045] The English practice module includes an English training unit and a self-assessment unit;

[0046] The English training unit is used for student users to practice through the first set of wrong questions and the second set of wrong questions;

[0047] The self-assessment unit is used for student users to perform simple assessments after repeated practice.

[0048] The present invention has the following beneficial effects:

[0049] 1. In the present invention, by setting up a data collection terminal, the data related to the English assessment is obtained and updated in real time, and the data is graded and classified according to the difficulty and type of the data. At the same time, the data is monitored to determine whether there are errors and duplications in the data. If errors and duplications occur, they can be corrected autonomously based on Internet data. If they cannot be corrected autonomously, an exception is entered and reported to the system for feedback, so that errors and duplications in the assessment data can be avoided during the English assessment, and the accuracy of the assessment results can be improved.

[0050] 2. In the present invention, a voice recognition terminal is set up, and a voice evaluation package is obtained by receiving the voice signal of the student user. At the same time, a video evaluation package is obtained by face-to-face communication between the virtual AI cartoon character and the student user. The voice signals in the voice evaluation package and the video evaluation package are recognized, and amplification, filtering and noise reduction processing are performed at the same time. The voice signal is automatically compared and judged with the data, and the judgment result and the corresponding score are automatically given. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still an objection, the abnormality can be appealed and reported to the system for feedback, making the English evaluation scene more realistic. The virtual AI cartoon character can simulate the real language communication environment and realize the face-to-face communication scene, so that the examinee's actual oral application ability can be better tested, thereby enhancing the practicality of this system.

[0051] 3. In the present invention, by setting up a memory practice terminal, after the automatic English assessment is completed, the judgment results are received through comparison and an electronic wrong question collection is formed. At the same time, the types of wrong questions are analyzed according to the electronic wrong question collection, and a large amount of materials of the same level and type are obtained from the materials according to the types of wrong questions, and the materials are stored. Student users can practice multiple times from the materials to master the difficult points of English, and can improve their own English ability through this system. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a schematic diagram of the overall system architecture of an automatic English assessment system based on speech recognition according to the present invention;

[0053] Figure 2 This is a schematic diagram of the architecture of a data collection terminal of an English automatic assessment system based on speech recognition according to the present invention;

[0054] Figure 3 This is a schematic diagram of the architecture of a speech recognition terminal of an English automatic assessment system based on speech recognition according to the present invention;

[0055] Figure 4 The figure is a schematic diagram of the architecture of the memory practice end of the automatic English assessment system based on speech recognition of the present invention. DETAILED DESCRIPTION

[0056] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0057] Example 1

[0058] Please refer to Figure 1 and Figure 2 As shown: An English automatic assessment system based on speech recognition includes a data collection end, a speech recognition end and a memory practice end. The data collection end and the speech recognition end are both provided with an abnormal feedback module;

[0059] The data collection terminal is used to obtain and update English assessment-related data in real time, and classify and process the data according to its difficulty and type. At the same time, the data is monitored to determine whether there are errors and duplications. If errors and duplications are found, they can be corrected automatically based on online data. If they cannot be corrected automatically, an exception will be generated and reported to the system for feedback.

[0060] The voice recognition end is used to receive the voice signal of the student user to obtain a voice evaluation package, and at the same time obtain a video evaluation package through face-to-face communication between the virtual AI cartoon character and the student user. The voice signal in the voice evaluation package and the video evaluation package is recognized, and amplified, filtered and noise-reduced. The voice signal is automatically compared and judged with the data, and the judgment result and corresponding score are automatically given. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still any objection, the student user can enter the complaint exception and report it to the system for feedback;

[0061] The memory practice terminal is used to receive the judgment results through comparison and form an electronic wrong question collection. At the same time, it analyzes the types of wrong questions based on the electronic wrong question collection, obtains a large amount of materials of the same level and type from the materials according to the types of wrong questions, and stores the materials. Student users can practice multiple times from the materials to master the difficult points of English.

[0062] The abnormality feedback module will provide feedback when an abnormality occurs in the system, and feed the abnormal information back to the technical staff to improve the system.

[0063] The data collection end includes a data acquisition module, a data monitoring module, and a data correction module;

[0064] The data acquisition module includes English assessment data unit, data update unit and grading and classification unit;

[0065] The English assessment data unit is used to store all relevant data used for English assessment, such as test papers, websites and books, through a storage device;

[0066] The data updating unit is used to update data in real time through the Internet;

[0067] The classification unit is used to classify materials according to their difficulty level and to classify them according to their type.

[0068] The data monitoring module includes a data comparison unit, a data duplication unit and a data judgment unit;

[0069] The data comparison unit is used to compare the acquired and real-time updated data with English grammar, vocabulary and sentences and obtain a first comparison result, which is divided into two categories according to errors and repetitions;

[0070] The data duplication unit is used to screen each assessment test question in the acquired and real-time updated data, and record the duplicate test questions as duplicate test questions;

[0071] The data judgment unit is used to reconfirm the first comparison result, upload the correct data to the data collection terminal, and record the erroneous and duplicated data as the second comparison result.

[0072] The data correction module includes a data search unit and an execution correction unit;

[0073] The data search unit is used to collect relevant content on the Internet based on the content of the second comparison result. If the second comparison result is a word, the pronunciation, letter composition and Chinese translation of the word are collected on the Internet. If the second comparison result is a sentence, the word composition, letter composition, Chinese translation and grammar of the sentence are collected on the Internet.

[0074] The execution correction unit is used to correct the repeated test questions and the second comparison result, delete the repeated test questions, and modify the second comparison result according to Internet information.

[0075] By acquiring and updating English assessment-related materials in real time, grading and classifying the materials according to their difficulty and type, and monitoring the materials to determine whether there are errors and duplications, if errors and duplications occur, they can be corrected autonomously based on Internet materials. If they cannot be corrected autonomously, they will be reported to the system for feedback, thus avoiding errors and duplications in the English assessment materials and improving the accuracy of the assessment results.

[0076] Example 2

[0077] Please refer to Figure 1 and 3 As shown: An English automatic assessment system based on speech recognition includes a data collection end, a speech recognition end and a memory practice end. The data collection end and the speech recognition end are both provided with an abnormal feedback module;

[0078] The data collection terminal is used to obtain and update English assessment-related data in real time, and classify and process the data according to its difficulty and type. At the same time, the data is monitored to determine whether there are errors and duplications. If errors and duplications are found, they can be corrected automatically based on online data. If they cannot be corrected automatically, an exception will be generated and reported to the system for feedback.

[0079] The voice recognition end is used to receive the voice signal of the student user to obtain a voice evaluation package, and at the same time obtain a video evaluation package through face-to-face communication between the virtual AI cartoon character and the student user. The voice signal in the voice evaluation package and the video evaluation package is recognized, and amplified, filtered and noise-reduced. The voice signal is automatically compared and judged with the data, and the judgment result and corresponding score are automatically given. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still any objection, the student user can enter the complaint exception and report it to the system for feedback;

[0080] The memory practice terminal is used to receive the judgment results through comparison and form an electronic wrong question collection. At the same time, it analyzes the types of wrong questions based on the electronic wrong question collection, obtains a large amount of materials of the same level and type from the materials according to the types of wrong questions, and stores the materials. Student users can practice multiple times from the materials to master the difficult points of English.

[0081] The abnormality feedback module will provide feedback when an abnormality occurs in the system, and feed the abnormal information back to the technical staff to improve the system.

[0082] The speech recognition end includes a speech receiving module, a virtual evaluation module, a speech processing module, and a judgment and scoring module;

[0083] The voice receiving module includes a voice receiving unit and a voice storage unit;

[0084] The voice receiving unit is used to receive the voice of the student user through the microphone;

[0085] The voice storage unit is used to save the student user's voice to obtain a voice evaluation package.

[0086] The virtual evaluation module includes a virtual communication unit and a video storage unit;

[0087] The virtual communication unit is used for the virtual AI cartoon character to conduct evaluation and communication with student users through the camera and generate videos;

[0088] The video storage unit is used to save the video to obtain a video evaluation package.

[0089] The speech processing module includes a speech signal extraction unit, a speech amplification unit and a speech filtering and noise reduction unit;

[0090] The voice signal extraction unit is used to identify the voice signals in the voice evaluation package and the video evaluation package and obtain the voice signals;

[0091] The voice amplification unit is used to amplify the sound of the voice signal;

[0092] The speech filtering and noise reduction unit is used to remove or weaken inaccurate signal frequencies and interference components in the speech signal, thereby obtaining cleaner and more accurate information.

[0093] The judgment and scoring module includes an evaluation comparison unit, an evaluation scoring unit, and an evaluation objection unit;

[0094] The comparison and evaluation unit compares the speech signal with the text and speech information in the data and displays the comparison results;

[0095] The evaluation and scoring unit is used to score the speech signal from 0 to 100 points based on the comparison results, obtain the scoring results and display the scoring information;

[0096] The assessment objection unit is used for student users to file an appeal when they disagree with the scoring results.

[0097] A voice evaluation package is obtained by receiving the voice signal of the student user, and a video evaluation package is obtained by face-to-face communication between the virtual AI cartoon character and the student user. The voice signals in the voice evaluation package and the video evaluation package are recognized, and amplified, filtered and noise reduced at the same time. The voice signal is automatically compared and judged with the data, and the judgment result and the corresponding score are automatically given. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still an objection, the abnormality can be appealed and reported to the system for feedback, making the English evaluation scene more realistic. The virtual AI cartoon character can simulate the real language communication environment and realize the face-to-face communication scene, so that the candidate's actual oral application ability can be better tested, thereby enhancing the practicality of this system.

[0098] Example 3

[0099] Please refer to Figure 1 and 4 As shown: An English automatic assessment system based on speech recognition includes a data collection end, a speech recognition end and a memory practice end. The data collection end and the speech recognition end are both provided with an abnormal feedback module;

[0100] The data collection terminal is used to obtain and update English assessment-related data in real time, and classify and process the data according to its difficulty and type. At the same time, the data is monitored to determine whether there are errors and duplications. If errors and duplications are found, they can be corrected automatically based on online data. If they cannot be corrected automatically, an exception will be generated and reported to the system for feedback.

[0101] The voice recognition end is used to receive the voice signal of the student user to obtain a voice evaluation package, and at the same time obtain a video evaluation package through face-to-face communication between the virtual AI cartoon character and the student user. The voice signal in the voice evaluation package and the video evaluation package is recognized, and amplified, filtered and noise-reduced. The voice signal is automatically compared and judged with the data, and the judgment result and corresponding score are automatically given. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still any objection, the student user can enter the complaint exception and report it to the system for feedback;

[0102] The memory practice terminal is used to receive the judgment results through comparison and form an electronic wrong question collection. At the same time, it analyzes the types of wrong questions based on the electronic wrong question collection, obtains a large amount of materials of the same level and type from the materials according to the types of wrong questions, and stores the materials. Student users can practice multiple times from the materials to master the difficult points of English.

[0103] The abnormality feedback module will provide feedback when an abnormality occurs in the system, and feed the abnormal information back to the technical staff to improve the system.

[0104] The memory practice end includes a wrong question acquisition module and an English practice module;

[0105] The wrong question acquisition module includes a wrong question receiving unit, a similar question acquisition unit and a wrong question collection unit;

[0106] The wrong question receiving unit is used to receive wrong questions in the comparison results and conduct a comprehensive analysis of the wrong questions;

[0107] The same type acquisition unit is used to search and obtain the same type of evaluation materials for correct and incorrect questions;

[0108] The wrong question collection unit is used to collect and store the wrong questions after each evaluation as a first wrong question set, and to collect and store the obtained evaluation materials of the same type as a second wrong question set.

[0109] The English practice module includes English training units and self-assessment units;

[0110] The English training unit is used for student users to practice through the first set of wrong questions and the second set of wrong questions;

[0111] The self-assessment unit is used for student users to conduct simple assessments after repeated practice.

[0112] After the automatic English assessment is completed, the judgment results are received through comparison and an electronic set of wrong questions is formed. At the same time, the types of wrong questions are analyzed based on the electronic set of wrong questions. A large amount of materials of the same level and type are obtained from the materials according to the types of wrong questions and stored. Student users can practice multiple times from the materials to master the difficult points of English, and can improve their own English ability through this system.

[0113] The present invention provides an automatic English assessment system based on speech recognition. When the system is in operation, it obtains and updates English assessment-related materials in real time, and classifies and processes the materials according to their difficulty and type. At the same time, it monitors the materials to determine whether there are errors and duplications in the materials. If errors and duplications occur, they can be corrected autonomously based on Internet materials. If they cannot be corrected autonomously, an exception is entered and reported to the system for feedback, so that errors and duplications in the assessment materials can be avoided during English assessment, and the accuracy of the assessment results can be improved. The system also receives the voice signals of student users to obtain voice assessment packages, and communicates face-to-face with student users through virtual AI cartoon characters to obtain video assessment packages. The system recognizes the voice signals in the voice assessment packages and video assessment packages, and performs amplification, filtering and noise reduction processing. The voice signals are automatically compared and judged with the materials, and automatically given The system will output the judgment results and corresponding scores. If the student user has any objection to the judgment result, the voice signal will be re-compared and judged. If there is still an objection, the user can enter the appeal exception and report the system for feedback, making the English assessment scene more realistic. It can simulate the real language communication environment through virtual AI cartoon characters and realize the face-to-face communication scene, so that the candidate's actual oral application ability can be better tested and the practicality of this system is enhanced. After the automatic English assessment is completed, the judgment results are received through comparison and an electronic wrong question set is formed. At the same time, the wrong question type is analyzed according to the electronic wrong question set. A large amount of the same level and type of information is obtained from the data according to the wrong question type and stored. Student users can practice multiple times from the data to master the English difficulties and can improve the student user's own English ability through this system. Finally, when the system has an abnormality, feedback is given and the abnormal information is fed back to the technical staff for system improvement.

[0114] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. An automatic English assessment system based on speech recognition, characterized in that: The English automatic assessment system based on speech recognition includes a data collection end, a speech recognition end and a memory practice end, wherein the data collection end and the speech recognition end are jointly provided with an abnormal feedback module; The data collection terminal is used to obtain and update data related to English assessment in real time, and classify and process the data according to the difficulty and type of the data. At the same time, the data is monitored to determine whether there are errors and duplications in the data. If errors and duplications are found, they can be corrected automatically based on the information on the Internet. If they cannot be corrected automatically, an exception is generated and reported to the system for feedback; The voice recognition terminal is used to receive the voice signal of the student user to obtain a voice evaluation package, and at the same time obtain a video evaluation package through face-to-face communication between the virtual AI cartoon character and the student user, recognize the voice signals in the voice evaluation package and the video evaluation package, and perform amplification, filtering and noise reduction processing at the same time, automatically compare and judge the voice signal with the data, and automatically give the judgment result and corresponding score. If the student user has any objection to the judgment result, the voice signal is re-compared and judged. If there is still an objection, the student user can enter the complaint exception and report the system for feedback; The memory practice terminal is used to receive the judgment results through comparison and form an electronic wrong question collection, and analyze the types of wrong questions according to the electronic wrong question collection. According to the wrong question types, a large amount of materials of the same level and type are obtained from the materials and stored. Students can practice multiple times from the materials to master the difficult points of English. The abnormality feedback module provides feedback when an abnormality occurs in the system, and feeds back the abnormality information to technical personnel for system improvement.

2. The system according to claim 1, wherein: The data collection terminal includes a data acquisition module, a data monitoring module and a data correction module; The data acquisition module includes an English assessment data unit, a data update unit and a grading and classification unit; The English assessment data unit is used to store all relevant data for English assessment such as test papers, websites and books through a storage device; The data updating unit is used to update data in real time via the Internet; The grading and classification unit is used to grade the data according to the degree of difficulty and classify the data according to the type of data.

3. The system according to claim 2, characterized in that The data monitoring module includes a data comparison unit, a data duplication unit and a data judgment unit; The data comparison unit is used to compare the acquired and real-time updated data with English grammar, vocabulary and sentences and obtain a first comparison result, which is divided into two categories according to errors and repetitions; The data duplication unit is used to screen each evaluation test question in the acquired and real-time updated data, and record the repeated test questions as repeated test questions; The data judgment unit is used to reconfirm the first comparison result, upload the correct data to the data collection terminal, and record the erroneous and repeated data as the second comparison result.

4. The system according to claim 3, characterized in that The data correction module includes a data search unit and a correction execution unit; The data search unit is used to collect relevant content on the Internet based on the content of the second comparison result. If the second comparison result is a word, the pronunciation, letter composition and Chinese translation of the word are collected on the Internet. If the second comparison result is a sentence, the word composition, letter composition, Chinese translation and grammar of the sentence are collected on the Internet. The execution correction unit is used to correct the repeated test questions and the second comparison result, delete the repeated test questions, and modify the second comparison result according to Internet information.

5. The system according to claim 1, wherein: The speech recognition terminal includes a speech receiving module, a virtual evaluation module, a speech processing module and a judgment and scoring module; The voice receiving module includes a voice receiving unit and a voice storage unit; The voice receiving unit is used to receive the voice of the student user through a microphone; The voice storage unit is used to save the voice of the student user to obtain a voice evaluation package.

6. The system according to claim 5, characterized in that The virtual evaluation module includes a virtual communication unit and a video storage unit; The virtual communication unit is used for the virtual AI cartoon character to conduct evaluation and communication with the student user through the camera and generate a video; The video storage unit is used to save the video to obtain a video evaluation package.

7. The system according to claim 6, characterized in that The speech processing module includes a speech signal extraction unit, a speech amplification unit and a speech filtering and noise reduction unit; The voice signal extraction unit is used to identify the voice signals in the voice evaluation package and the video evaluation package and obtain the voice signals; The voice amplification unit is used to amplify the sound of the voice signal; The speech filtering and noise reduction unit is used to remove or weaken inaccurate signal frequencies and interference components in the speech signal, thereby obtaining cleaner and more accurate information.

8. The system according to claim 7, characterized in that The judgment and scoring module includes an evaluation and comparison unit, an evaluation and scoring unit, and an evaluation objection unit; The comparison and evaluation unit compares the voice signal with the text and voice information in the data and displays the comparison result; The evaluation and scoring unit is used to score the speech signal from 0 to 100 points based on the comparison result, obtain the scoring result and display the scoring information; The evaluation objection unit is used for students to file an appeal when they have objections to the scoring results.

9. The system according to claim 1, wherein: The memory practice terminal includes a wrong question acquisition module and an English practice module; The wrong question acquisition module includes a wrong question receiving unit, a similar question acquisition unit and a wrong question collection unit; The wrong question receiving unit is used to receive the wrong questions in the comparison result and perform a comprehensive analysis on the wrong questions; The same type acquisition unit is used to search and acquire the same type of evaluation materials for wrong questions; The wrong question collection unit is used to collect and store the wrong questions after each evaluation as a first wrong question collection, and to collect and store the obtained evaluation materials of the same type as a second wrong question collection.

10. The system according to claim 9, characterized in that The English practice module includes an English training unit and a self-assessment unit; The English training unit is used for student users to practice through the first set of wrong questions and the second set of wrong questions; The self-assessment unit is used for student users to perform simple assessments after repeated practice.

Images