A teaching quality assessment method and system based on big data

Abstract

This invention relates to the field of intelligent assessment technology, specifically disclosing a teaching level assessment method and system based on big data. The method includes recording classroom interaction information using intelligent devices, identifying the classroom interaction information, and extracting dialogue information; obtaining a student score array containing time nodes, arranging it according to time order to obtain a student score matrix; comparing dialogue information from different classrooms, clustering the classrooms, reading the student score matrices of similar classrooms, and internally correcting the student score matrices of similar classrooms; receiving student-input rating information, statistically analyzing the rating information and the corrected student score matrix to determine the total assessment score. This invention corrects the student score matrix based on dialogue information as an assessment parameter, ensuring strong objectivity. Combined with existing subjective evaluation processes, it enriches the evaluation process by integrating subjective and objective elements.

Description

Technical Field

[0001] This invention relates to the field of intelligent assessment technology, specifically a teaching level assessment method and system based on big data. Background Technology

[0002] Teaching quality assessment is a process of systematically evaluating the teaching ability, teaching effectiveness, and professional competence demonstrated by instructors during the teaching process. Existing teaching quality assessment methods are relatively simplistic, relying on students to score the lessons after they conclude, resulting in an assessment. This method is overly subjective and lacks objectivity. In reality, with the widespread use of smart devices in classrooms, more objective evaluation methods can be introduced into the teaching process. How to use smart devices to objectively evaluate the teaching process is the technical problem this invention aims to solve. Summary of the Invention

[0003] The purpose of this invention is to provide a teaching level assessment method and system based on big data to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A teaching quality assessment method based on big data, the method comprising:

[0006] Classroom interaction information is recorded using smart devices, and the classroom interaction information is identified and dialogue information is extracted; wherein, the classroom interaction information includes at least audio information; the classroom interaction information and the extracted dialogue information have the same time period label;

[0007] Obtain the student score array containing time nodes, arrange it according to time order to obtain the student score matrix; the rows of the student score matrix correspond to time nodes, the columns correspond to students, and the element values ​​in the matrix represent scores;

[0008] By comparing the dialogue information of different classes, the classes are clustered, the student score matrix of the same type of class is read, and the student score matrix of the same type of class is internally corrected; the internal correction method is to reduce the difference.

[0009] Receive the rating information input by the students, and use the rating information and the corrected student score matrix to determine the total evaluation score.

[0010] As a further aspect of the present invention: the step of recording classroom interaction information based on a smart device, identifying the classroom interaction information, and extracting dialogue information includes:

[0011] Real-time recording of classroom interaction information using smart devices;

[0012] When the classroom interaction information is audio, the teacher's audio characteristics are identified based on preset keywords;

[0013] The audio is segmented based on the aforementioned audio features to obtain teacher audio and non-teacher audio; both teacher audio and non-teacher audio contain timestamps.

[0014] Audio recognition is performed on both teacher and non-teacher audio to obtain dialogue information.

[0015] As a further aspect of the present invention: the steps of comparing dialogue information from different classrooms, clustering the classrooms, reading the student score matrix of similar classrooms, and internally correcting the student score matrix of similar classrooms include:

[0016] For any classroom dialogue information, word recognition is performed, and the dialogue information is simplified into a sequence of keyword groups based on the word recognition results; each keyword group corresponds to a dialogue segment.

[0017] Input the sequence of keyword groups into the preset conversion model to obtain the sequence of word vector groups;

[0018] Compare the word vector sequences of different classrooms, calculate the similarity, and cluster the classrooms based on the similarity.

[0019] Read the student score matrix of similar classes and internally correct it.

[0020] As a further aspect of the present invention: the step of receiving the scoring information input by the student, statistically analyzing the scoring information and the corrected student score matrix to determine the total evaluation score includes:

[0021] Receive rating information input by trainees;

[0022] The scoring information and the corrected student score matrix are input into the trained evaluation model to obtain the total evaluation score; wherein, the evaluation model is a preset mathematical statistical function.

[0023] As a further aspect of the present invention, the method further includes:

[0024] For any class's student score matrix, calculate the mean and standard deviation of each row of data to obtain the mean column and standard deviation column;

[0025] By comparing the mean and standard deviation columns of different classrooms, the similarity between different classrooms is calculated, and the classrooms are clustered based on the similarity.

[0026] For any type of classroom, read the dialogue information of each classroom, compare the dialogue information, extract the common dialogue information, and use it as the standard dialogue information for that type of classroom.

[0027] The modal dialogue information is used to characterize the modal features of dialogue information in similar classrooms.

[0028] As a further aspect of the present invention, the method further includes:

[0029] Receive previously acquired dialogue information uploaded from a specific classroom;

[0030] The acquired dialogue information is regularly compared and matched with existing dialogue information to obtain a matching classroom.

[0031] Query the student score matrix of the matching class and select the optimal student score matrix;

[0032] Read the standard dialogue information corresponding to the optimal student score matrix as reference data and feed it back to the classroom that uploaded the dialogue information.

[0033] The present invention also provides a teaching level assessment system based on big data, the system comprising:

[0034] The dialogue information extraction module is used to record classroom interaction information based on smart devices, identify the classroom interaction information, and extract dialogue information; wherein, the classroom interaction information includes at least audio information; the classroom interaction information and the extracted dialogue information have the same time period label;

[0035] The matrix generation module is used to obtain an array of student scores containing time nodes, arrange them in chronological order, and obtain a student score matrix; the rows of the student score matrix correspond to time nodes, the columns correspond to students, and the element values ​​in the matrix represent scores;

[0036] The matrix correction module is used to compare dialogue information from different classrooms, cluster classrooms, read student score matrices from classrooms of the same type, and internally correct the student score matrices from classrooms of the same type; the internal correction method is to reduce the differences.

[0037] The total assessment score generation module is used to receive the scoring information input by the students, and to determine the total assessment score by statistically analyzing the scoring information and the corrected student score matrix.

[0038] As a further aspect of the present invention: the dialogue information extraction module includes:

[0039] The information recording unit is used to record classroom interaction information in real time using smart devices;

[0040] An audio feature extraction unit is used to identify the teacher's audio features based on preset keywords when the classroom interaction information is audio.

[0041] An audio segmentation unit is used to segment the audio according to the audio features to obtain teacher audio and non-teacher audio; wherein both teacher audio and non-teacher audio contain timestamps;

[0042] The audio recognition unit is used to recognize audio from both the teacher's and non-teacher's audio to obtain dialogue information.

[0043] As a further aspect of the present invention: the matrix correction module includes:

[0044] The information conversion unit is used to perform word recognition on dialogue information in any classroom, and simplify the dialogue information into a sequence of keyword groups based on the word recognition results; each keyword group corresponds to a dialogue segment.

[0045] The word conversion unit is used to input the sequence of keyword groups into a preset conversion model to obtain a sequence of word vector groups;

[0046] The comparison clustering unit is used to compare the word vector groups of different classrooms, calculate the similarity, and cluster the classrooms based on the similarity.

[0047] The correction execution unit is used to read the student score matrix of similar classes and internally correct the student score matrix of similar classes.

[0048] As a further aspect of the present invention: the total score generation module includes:

[0049] The rating information receiving unit is used to receive the rating information input by the trainees;

[0050] An evaluation execution unit is used to input the scoring information and the corrected student score matrix into a trained evaluation model to obtain a total evaluation score; wherein, the evaluation model is a preset mathematical statistical function.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention records classroom interaction information based on intelligent devices, determines dialogue information based on classroom interaction information, synchronously obtains students' scores at each moment, obtains a student score matrix, and corrects the student score matrix based on dialogue information as an evaluation parameter. It is highly objective and combines the existing subjective evaluation process, thus enriching the evaluation process by combining subjective and objective factors. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0053] Figure 1 This is a flowchart of a big data-based teaching quality assessment method.

[0054] Figure 2 This is a structural diagram of a teaching quality assessment system based on big data. Detailed Implementation

[0055] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0056] Figure 1 This is a flowchart of a teaching quality assessment method based on big data. In this embodiment of the invention, a teaching quality assessment method based on big data includes:

[0057] Step S100: Record classroom interaction information based on a smart device, identify the classroom interaction information, and extract dialogue information; wherein, the classroom interaction information includes at least audio information; the classroom interaction information and the extracted dialogue information have the same time period label;

[0058] The technical solution of this invention is applied to smart classrooms, which are classrooms equipped with smart devices and are very common in existing teaching environments. Many students are equipped with tablets during class, and the process of recording interactive information is very simple. In the technical solution of this invention, the recorded interactive information is limited to audio information and generally does not involve image information. On the one hand, image information has a large amount of data, and on the other hand, people may be more resistant to image information and will not easily grant authorization. By recognizing the classroom interactive information, dialogue information can be extracted.

[0059] It should be noted that even when acquiring audio information, it is necessary to obtain permission from both the instructor and the student beforehand. Without the necessary permissions, subsequent processing is not possible.

[0060] Step S200: Obtain the student score array containing time nodes, arrange it according to time order to obtain the student score matrix; the rows of the student score matrix correspond to time nodes, the columns correspond to students, and the element values ​​in the matrix represent scores;

[0061] Obtain the student scores for each exam, and statistically analyze them according to the same sequence number to obtain an array of student scores for each time point. Sort the arrays of student scores in chronological order, and then concatenate them into a matrix called the student score matrix.

[0062] It should be noted that the data acquisition process in steps S100 and S200 is the final statistical process after a teaching process is completed. For example, if a teaching session consists of 36 class hours, the classroom interaction information is all the classroom interaction information within these 36 class hours, and the student score matrix is ​​the scores of all the exams within these 36 class hours.

[0063] Step S300: Compare the dialogue information of different classrooms, cluster the classrooms, read the student score matrix of the same type of classroom, and internally correct the student score matrix of the same type of classroom; wherein, the internal correction method is to reduce the difference.

[0064] The data statistics process in steps S100 and S200 involves multiple classrooms and is itself a big data acquisition architecture. The number of classrooms is extremely large. By comparing the dialogue information of different classrooms, the classrooms can be clustered based on the comparison results, grouping classrooms with similar teaching processes into one category. Then, the student score matrix of the same category of classrooms is read and internally corrected. The goal of this internal correction is to reduce the difference between the best and worst student score matrices. Reducing this difference aims to minimize the differences in teaching effectiveness caused by student differences. In fact, calculating the mean of the student score matrix of the same category of classrooms as the final evaluation of teaching results is feasible, but this is too balanced and erases the differences among instructors, including subtle differences in facial expressions, demeanor, and posture. Therefore, this application only reduces the differences, not completely eliminates them.

[0065] Step S400: Receive the rating information input by the student, and determine the total evaluation score by summarizing the rating information and the corrected student score matrix;

[0066] The existing evaluation system involves anonymous scoring by students. This method is highly subjective and easily influenced by interference, but it is a necessary scoring mechanism, so it is retained in this invention. In the technical solution of this invention, the generated student score matrix actually represents objective information. Combining subjective and objective aspects to evaluate the teaching process is more accurate.

[0067] Regarding step S100, the step of recording classroom interaction information based on smart devices, identifying the classroom interaction information, and extracting dialogue information includes:

[0068] Real-time recording of classroom interaction information using smart devices;

[0069] When the classroom interaction information is audio, the teacher's audio characteristics are identified based on preset keywords;

[0070] The audio is segmented based on the aforementioned audio features to obtain teacher audio and non-teacher audio; both teacher audio and non-teacher audio contain timestamps.

[0071] Audio recognition is performed on both teacher and non-teacher audio to obtain dialogue information.

[0072] The above provides a specific scheme for extracting dialogue information. Based on real-time recording of classroom interaction information by smart devices, when the classroom interaction information is audio, the teacher's audio characteristics are identified according to preset keywords. There are many keywords, such as "students," which is an audio phrase frequently uttered by the teacher. These keywords are identified, and the frequency component is extracted to extract audio features (equivalent to identifying the teacher's timbre). The audio (classroom interaction information) is segmented according to the audio features to obtain the teacher's audio and non-teacher's audio. Finally, audio recognition is performed on the teacher's audio and non-teacher's audio to obtain the dialogue information. There are many existing recognition schemes available, and this is a simple audio-to-text process, which will not be elaborated further.

[0073] It is worth noting that this application does not identify trainees, but only instructors; all other personnel are referred to as non-instructors. In addition, both instructor audio and non-instructor audio contain timestamps to determine the order of the identified text.

[0074] Regarding step S300, the steps of comparing dialogue information from different classrooms, clustering classrooms, reading student score matrices from similar classrooms, and internally correcting student score matrices from similar classrooms include:

[0075] For any classroom dialogue information, word recognition is performed, and the dialogue information is simplified into a sequence of keyword groups based on the word recognition results; each keyword group corresponds to a dialogue segment.

[0076] Input the sequence of keyword groups into the preset conversion model to obtain the sequence of word vector groups;

[0077] Compare the word vector sequences of different classrooms, calculate the similarity, and cluster the classrooms based on the similarity.

[0078] Read the student score matrix of similar classes and internally correct the student score matrix of similar classes;

[0079] In one example of the technical solution of this invention, the dialogue information of each classroom is identified sequentially. For the dialogue information of any classroom, word recognition is performed to convert the dialogue information into a set of words. Each sentence in the dialogue information can be converted into a keyword group. Therefore, the entire dialogue information can be converted into a sequence of keyword groups. The order in the sequence is the time order determined by the timestamp.

[0080] Then, the sequence of keyword groups is input into a preset conversion model, each word is converted into a word vector, the string is quantized, and a sequence of word vector groups is obtained; the sequence of word vector groups of different classrooms is compared, the similarity is calculated, and the classrooms are clustered based on the similarity to obtain multiple types of classrooms; the clustering scheme can adopt the K-means clustering scheme, the K value can be a preset value, generally not unique, multiple K values ​​can be applied, multiple clustering is performed, and then the best K value is selected.

[0081] Finally, for each type of class, the student score matrix of each class is read, and any student score matrix is ​​corrected according to the student score matrix of the same type of class to obtain the final student score matrix.

[0082] The internal correction method is as follows:

[0083] Calculate the mean value at each element position in the student score matrix;

[0084] For any element position in any student's score matrix, calculate the mean of the element value and the corresponding mean, and use it as the corrected data;

[0085] When a student's score matrix does not contain a corresponding value at a certain element position, the original value is retained during the internal correction process.

[0086] The above provides a specific internal correction scheme. First, the score matrix of all students is statistically analyzed, and the mean of the values ​​at each element position is calculated. Then, for any value in the student score matrix, the mean of its value and the corresponding mean is calculated, which is used as the corrected value. The corrected value is closer to the mean. In fact, during the correction, a weight can be added. If we want the original value to have a greater influence, we can make its weight two-thirds; if we want the original value to have a smaller influence, we can make its weight one-third.

[0087] Finally, it should be noted that since the size of the student score matrix may be different, some positions may only have values ​​in the student score matrix. This application does not calculate such values, but simply retains the original values.

[0088] Regarding step S400, the step of receiving the rating information input by the student, statistically analyzing the rating information and the corrected student score matrix to determine the total evaluation score includes:

[0089] Receive rating information input by trainees;

[0090] The scoring information and the corrected student score matrix are input into the trained evaluation model to obtain the total evaluation score; wherein, the evaluation model is a preset mathematical statistical function.

[0091] In one example of the technical solution of the present invention, the scoring information and the corrected student score matrix are input into a trained evaluation model to obtain the total evaluation score. The evaluation model is a preset mathematical statistical function, which is determined by the staff as needed. For example, the mean of the scoring information is calculated, and the average improvement score of the students is calculated based on the student score matrix (the difference between the maximum and minimum values ​​in any column, and then the mean of the differences in all columns is calculated). The mean and the average improvement score are input into a preset linear function in two variables to calculate the total score.

[0092] As a preferred embodiment of the technical solution of the present invention, the method further includes:

[0093] For any class's student score matrix, calculate the mean and standard deviation of each row of data to obtain the mean column and standard deviation column;

[0094] By comparing the mean and standard deviation of different classes, the similarity between the classes is calculated, and the classes are clustered based on the similarity.

[0095] For any type of classroom, read the dialogue information of each classroom, compare the dialogue information, and extract the common dialogue information as the standard dialogue information for that type of classroom.

[0096] For any given classroom's student score matrix, feature extraction is performed to obtain the mean and standard deviation columns. The mean and standard deviation columns of different classrooms are compared to calculate their similarity (using an array similarity calculation scheme). Based on this similarity, classrooms are clustered, grouping those with similar score changes into one category. Further clustering is then performed on these similar classrooms, based on dialogue information. This reveals the dialogue information under similar score changes. The class with the most classes is selected, and its dialogue information is used as the standard dialogue information, serving as the dependent variable for the student score matrix. The practical significance of this is that if teaching is conducted according to the standard dialogue information, a corresponding student score matrix is ​​highly likely to be obtained.

[0097] In summary, the modal dialogue information is used to characterize the modal features of dialogue information in similar classrooms.

[0098] As a preferred embodiment of the technical solution of the present invention, the method further includes:

[0099] Receive previously acquired dialogue information uploaded from a specific classroom;

[0100] The acquired dialogue information is regularly compared and matched with existing dialogue information to obtain a matching classroom.

[0101] Query the student score matrix of the matching class and select the optimal student score matrix;

[0102] Read the standard dialogue information corresponding to the optimal student score matrix as reference data and feed it back to the classroom that uploaded the dialogue information.

[0103] In this invention, the evaluation of teaching level is conducted after the teaching is completed and the final state is reached. This evaluation is used to intelligently score the instructors. In fact, this information can also serve as a reference during the teaching process. The system receives dialogue information uploaded from a classroom and periodically compares and matches it with existing dialogue information (extracting dialogue information of the same length from existing dialogue information and calculating the similarity; a sufficiently high similarity indicates a successful match). This results in a matched classroom. "Periodic" can mean every two class periods. There may be many matched classrooms. The system queries the student score matrix of the matched classrooms and selects the optimal student score matrix (with the highest average improvement score), considering it the best state achieved from the existing data. Then, based on the standard dialogue information corresponding to each student score matrix, the system queries the standard dialogue information corresponding to the optimal student score matrix as a teaching reference. This reference is fed back to the classroom that uploaded the dialogue information as guidance, assisting its teaching process.

[0104] Figure 2 This is a structural diagram of a teaching level assessment system based on big data. In this embodiment of the invention, a teaching level assessment system based on big data, system 10 includes:

[0105] The dialogue information extraction module 11 is used to record classroom interaction information based on a smart device, identify the classroom interaction information, and extract dialogue information; wherein, the classroom interaction information includes at least audio information; the classroom interaction information and the extracted dialogue information have the same time period label;

[0106] The matrix generation module 12 is used to obtain an array of student scores containing time nodes, arrange them in chronological order, and obtain a student score matrix; the rows of the student score matrix correspond to time nodes, the columns correspond to students, and the element values ​​in the matrix represent scores;

[0107] The matrix correction module 13 is used to compare the dialogue information of different classrooms, cluster the classrooms, read the student score matrix of the same type of classroom, and internally correct the student score matrix of the same type of classroom; wherein, the internal correction method is to reduce the difference.

[0108] The total evaluation score generation module 14 is used to receive the scoring information input by the students, and to determine the total evaluation score by statistically analyzing the scoring information and the corrected student score matrix.

[0109] Furthermore, the dialogue information extraction module 11 includes:

[0110] The information recording unit is used to record classroom interaction information in real time using smart devices;

[0111] An audio feature extraction unit is used to identify the teacher's audio features based on preset keywords when the classroom interaction information is audio.

[0112] An audio segmentation unit is used to segment the audio according to the audio features to obtain teacher audio and non-teacher audio; wherein both teacher audio and non-teacher audio contain timestamps;

[0113] The audio recognition unit is used to recognize audio from both the teacher's and non-teacher's audio to obtain dialogue information.

[0114] Specifically, the matrix correction module 12 includes:

[0115] The information conversion unit is used to perform word recognition on dialogue information in any classroom, and simplify the dialogue information into a sequence of keyword groups based on the word recognition results; each keyword group corresponds to a dialogue segment.

[0116] The word conversion unit is used to input the sequence of keyword groups into a preset conversion model to obtain a sequence of word vector groups;

[0117] The comparison clustering unit is used to compare the word vector groups of different classrooms, calculate the similarity, and cluster the classrooms based on the similarity.

[0118] The correction execution unit is used to read the student score matrix of similar classes and internally correct the student score matrix of similar classes.

[0119] In addition, the total score generation module 14 includes:

[0120] The rating information receiving unit is used to receive the rating information input by the trainees;

[0121] An evaluation execution unit is used to input the scoring information and the corrected student score matrix into a trained evaluation model to obtain a total evaluation score; wherein, the evaluation model is a preset mathematical statistical function.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A teaching level assessment method based on big data, characterized in that, The method includes: Classroom interaction information is recorded using smart devices, and the classroom interaction information is identified and dialogue information is extracted; wherein, the classroom interaction information includes at least audio information; the classroom interaction information and the extracted dialogue information have the same time period label; Obtain the student score array containing time nodes, arrange it according to time order to obtain the student score matrix; the rows of the student score matrix correspond to time nodes, the columns correspond to students, and the element values ​​in the matrix represent scores; By comparing the dialogue information of different classes, the classes are clustered, the student score matrix of the same type of class is read, and the student score matrix of the same type of class is internally corrected; the internal correction method is to reduce the difference. Receive the rating information input by the students, and determine the total evaluation score by summarizing the rating information and the corrected student score matrix. The steps of comparing dialogue information from different classrooms, clustering classrooms, reading student score matrices from similar classrooms, and internally correcting student score matrices from similar classrooms include: For any classroom dialogue information, word recognition is performed, and the dialogue information is simplified into a sequence of keyword groups based on the word recognition results; each keyword group corresponds to a dialogue segment. Input the sequence of keyword groups into the preset conversion model to obtain the sequence of word vector groups; Compare the word vector sequences of different classrooms, calculate the similarity, and cluster the classrooms based on the similarity. Read the student score matrix of similar classes and internally correct the student score matrix of similar classes; The internal correction method is as follows: calculate the mean value at each element position in the student score matrix; for any element position in any student score matrix, calculate the mean value of the element value and the corresponding mean value, and use it as the corrected data; when there is a student score matrix that does not contain a corresponding value at a certain element position, the original value is retained during the internal correction process.

2. The teaching level assessment method based on big data according to claim 1, characterized in that, The steps of recording classroom interaction information based on smart devices, identifying the classroom interaction information, and extracting dialogue information include: Real-time recording of classroom interaction information using smart devices; When the classroom interaction information is audio, the teacher's audio characteristics are identified based on preset keywords; The audio is segmented based on the aforementioned audio features to obtain teacher audio and non-teacher audio; both teacher audio and non-teacher audio contain timestamps. Audio recognition is performed on both teacher and non-teacher audio to obtain dialogue information.

3. The teaching level assessment method based on big data according to claim 1, characterized in that, The steps of receiving the rating information input by the students, statistically analyzing the rating information and the corrected student score matrix to determine the total evaluation score include: Receive rating information input by trainees; The scoring information and the corrected student score matrix are input into the trained evaluation model to obtain the total evaluation score; wherein, the evaluation model is a preset mathematical statistical function.

4. The teaching level assessment method based on big data according to claim 1, characterized in that, The method further includes: For any class's student score matrix, calculate the mean and standard deviation of each row of data to obtain the mean column and standard deviation column; By comparing the mean and standard deviation columns of different classrooms, the similarity between different classrooms is calculated, and the classrooms are clustered based on the similarity. For any type of classroom, read the dialogue information of each classroom, compare the dialogue information, extract the common dialogue information, and use it as the standard dialogue information for that type of classroom. The modal dialogue information is used to characterize the modal features of dialogue information in similar classrooms.

5. The teaching level assessment method based on big data according to claim 4, characterized in that, The method further includes: Receive previously acquired dialogue information uploaded from a specific classroom; The acquired dialogue information is regularly compared and matched with existing dialogue information to obtain a matching classroom. Query the student score matrix of the matching class and select the optimal student score matrix; Read the standard dialogue information corresponding to the optimal student score matrix as reference data and feed it back to the classroom that uploaded the dialogue information.

6. A teaching level assessment system based on big data, characterized in that, The system includes: The dialogue information extraction module is used to record classroom interaction information based on smart devices, identify the classroom interaction information, and extract dialogue information; wherein, the classroom interaction information includes at least audio information; the classroom interaction information and the extracted dialogue information have the same time period label; The matrix generation module is used to obtain an array of student scores containing time nodes, arrange them in chronological order, and obtain a student score matrix; the rows of the student score matrix correspond to time nodes, the columns correspond to students, and the element values ​​in the matrix represent scores; The matrix correction module is used to compare dialogue information from different classrooms, cluster classrooms, read student score matrices from classrooms of the same type, and internally correct the student score matrices from classrooms of the same type; the internal correction method is to reduce the differences. The total evaluation score generation module is used to receive the rating information input by the students, and to determine the total evaluation score by statistically analyzing the rating information and the corrected student score matrix. The matrix correction module includes: The information conversion unit is used to perform word recognition on dialogue information in any classroom, and simplify the dialogue information into a sequence of keyword groups based on the word recognition results; each keyword group corresponds to a dialogue segment. The word conversion unit is used to input the sequence of keyword groups into a preset conversion model to obtain a sequence of word vector groups; The comparison clustering unit is used to compare the word vector groups of different classrooms, calculate the similarity, and cluster the classrooms based on the similarity. The correction execution unit is used to read the student score matrix of similar classes and internally correct the student score matrix of similar classes. The internal correction method is as follows: calculate the mean value at each element position in the student score matrix; for any element position in any student score matrix, calculate the mean value of the element value and the corresponding mean value, and use it as the corrected data; when there is a student score matrix that does not contain a corresponding value at a certain element position, the original value is retained during the internal correction process.

7. The teaching level assessment system based on big data according to claim 6, characterized in that, The dialogue information extraction module includes: The information recording unit is used to record classroom interaction information in real time using smart devices; An audio feature extraction unit is used to identify the teacher's audio features based on preset keywords when the classroom interaction information is audio. An audio segmentation unit is used to segment the audio according to the audio features to obtain teacher audio and non-teacher audio; wherein both teacher audio and non-teacher audio contain timestamps; The audio recognition unit is used to recognize audio from both the teacher's and non-teacher's audio to obtain dialogue information.

8. The teaching level assessment system based on big data according to claim 6, characterized in that, The total assessment score generation module includes: The rating information receiving unit is used to receive the rating information input by the trainees; An evaluation execution unit is used to input the scoring information and the corrected student score matrix into a trained evaluation model to obtain a total evaluation score; wherein, the evaluation model is a preset mathematical statistical function.

Images