Class learning condition analysis system based on intelligent correction system

Abstract

The invention relates to a class learning condition analysis system based on an intelligent correction system, which is characterized in that a data aggregation device, a diagnosis device and a stratified analysis device are arranged on the basis of the existing intelligent correction system, and correction data of all students in a class are obtained through the data aggregation device; knowledge mastering feature extraction, error mode feature extraction and capability feature extraction are carried out on the corrected data through a diagnosis device to obtain class learning features, and then class core weak point evaluation based on a weighted scoring method is carried out on the class learning features to obtain class common features. All students in a class are divided into a plurality of sub-groups according to class generality characteristics through a hierarchical analysis device, and corresponding promotion direction suggestions, adaptive resource suggestions and teaching strategy suggestions are generated for weak generality of the students in each sub-group, so that the situation that a one-step strategy is adopted in a teaching decision can be effectively avoided, the requirements of different sub-groups are met, and the teaching efficiency is improved. And an educator can timely master the teaching effect.

Description

Technical Field

[0001] This invention relates to the field of interdisciplinary technology of artificial intelligence and education, and more specifically, to a class learning analysis system based on an intelligent grading system. Background Technology

[0002] Classroom learning analysis refers to the analysis of learning data from students in a designated class to diagnose the overall knowledge acquisition, skill gaps, and learning habit deficiencies of the students. Classroom learning analysis is a crucial basis for teachers to adjust teaching strategies and achieve precise teaching. Traditional classroom learning analysis relies on teachers manually analyzing homework corrections, exam results, and student classroom performance over a sustained period. Besides being time-consuming and labor-intensive, this method often overlooks fine-grained information in the corrections, including error types, distribution of incorrect answers, and assessments of knowledge connections, due to some teachers' insufficient statistical analysis skills.

[0003] In response, existing technologies have disclosed a type of intelligent grading system. This type of intelligent grading system grades the class's regular homework and periodic test questions to calculate the accuracy rate of each question and the questions with low accuracy rates. Based on the score range, the students in the class are divided into several levels, such as excellent, average and weak. Finally, the data formed by the level, the number of wrong questions, and the questions with low accuracy rates are compiled into data reports for educators to refer to.

[0004] However, these intelligent grading systems focus primarily on the accuracy of answers while neglecting the assessment of error types and knowledge connection abilities. This makes it difficult for educators to assess teaching effectiveness in a timely manner and thus hinders targeted lesson preparation. Furthermore, because these systems typically categorize students based on their scores, educators often adopt a one-size-fits-all approach in their teaching decisions, failing to meet the needs of different subgroups. Summary of the Invention

[0005] The technical problem to be solved by the present invention is how to overcome the technical defects of existing intelligent grading systems, which make it difficult for educators to prepare lessons in a targeted manner because they cannot grasp the teaching effect in a timely manner, and also fail to meet the needs of different subgroups. In order to overcome the above-mentioned defects of the prior art, the present invention provides a class learning situation analysis system based on an intelligent grading system.

[0006] This invention provides a class learning analysis system based on an intelligent grading system, comprising an intelligent grading system with project-based grading and recording functions, and further comprising: A data aggregation device is electrically connected to the output end of the intelligent grading system, and is used to collect the output data of the intelligent grading system in batches and integrate the output data to obtain the grading data of all students in the class. The diagnostic device, electrically connected to the data aggregation device, is used to extract knowledge mastery features, error pattern features, and ability features from the graded data to obtain class learning features. Then, the class learning features are used to evaluate the class's core weaknesses based on a weighted scoring method to obtain common features of the class. The hierarchical analysis device, electrically connected to the diagnostic device, is used to divide all students in the class into multiple subgroups based on the common characteristics of the class, and to generate corresponding suggestions for improvement, appropriate resources, and teaching strategies for the common weaknesses of students in each subgroup.

[0007] The class learning analysis system based on an intelligent grading system disclosed in this invention addresses the technical problems of this invention by incorporating a data aggregation device, a diagnostic device, and a hierarchical analysis device into existing intelligent grading systems. The data aggregation device obtains the grading data of all students in the class, and the diagnostic device extracts knowledge mastery features, error pattern features, and ability features from the grading data to obtain class learning characteristics. These characteristics are then used to assess the class's core weaknesses based on a weighted scoring method to obtain common class characteristics. Because this process utilizes knowledge mastery features, error pattern features, and ability characteristics, it not only focuses on the accuracy of questions but also reflects the assessment of error types and knowledge-related abilities, ensuring that educators can promptly grasp teaching effectiveness and prepare lessons accordingly. Simultaneously, the hierarchical analysis device divides all students in the class into multiple subgroups based on the common class characteristics and generates corresponding suggestions for improvement, appropriate resources, and teaching strategies for each subgroup based on their common weaknesses. This process categorizes students based on common class characteristics constructed from knowledge mastery features, error pattern features, and ability characteristics, rather than simple score ranges, effectively avoiding a one-size-fits-all approach in teaching decisions. Furthermore, after subgrouping, suggestions for improvement, appropriate resources, and teaching strategies are generated based on the common weaknesses of students in each subgroup, effectively realizing the educational loop and improving the educational process, thereby meeting the needs of different subgroups.

[0008] In one possible implementation, the data aggregation device includes: The grading data acquisition module is electrically connected to the output terminal of the intelligent grading system and is used to collect the output data in batches. The output data includes question data, student answer data, and basic data containing the number of students, subject, class number, and grading time. The data preprocessing module is electrically connected to both the graded data acquisition module and the diagnostic device. It is used to sequentially perform invalid data filtering, dimension alignment, and aggregation calculation on the output data to obtain the graded data of all students in the class.

[0009] A data aggregation device with the aforementioned structure and functions, based on the acquired output data, sequentially performs invalid data filtering, dimension alignment, and aggregation calculations. This removes abnormal or meaningless data, eliminates noise, and, through dimension alignment, facilitates subsequent feature calculations, improving computational efficiency. Data aggregation can then generate class statistical indicators, providing convenience for subsequent statistical inferences.

[0010] In one possible implementation, the class learning characteristics include each student's accuracy rate for questions corresponding to each knowledge point, the standard deviation of each student's accuracy rate, each student's accuracy rate in analyzing comprehensive questions across different knowledge points, each student's error type encoding feature vector, and each student's ability dimension feature vector. The calculation or collection of this data not only focuses on question accuracy but also specifically on error types and the assessment of knowledge association abilities, significantly improving the timeliness and comprehensiveness of the data and helping educators to promptly grasp teaching effectiveness.

[0011] In one possible implementation, the diagnostic device includes: The learning feature extraction module communicates with the data preprocessing module to extract knowledge mastery features from the graded data, thereby obtaining the accuracy rate of each student for each knowledge point, the standard deviation of each student's accuracy rate, and the accuracy rate of each student in analyzing comprehensive questions across knowledge points. Subsequently, error pattern features are extracted from the graded data to obtain the error type encoding feature vector for each student. Finally, ability features are extracted from the graded data to obtain the ability dimension feature vector for each student. The problem diagnosis module is electrically connected to both the learning feature extraction module and the hierarchical analysis device. It is used to calculate the Hadamard product of the error type encoding feature vector and the weight vector of each student to obtain the error vector of each student. Then, it calculates the mode of each component of the error vector and the ability dimension feature vector, and integrates the modes of all components of the error vector and the ability dimension feature vector into a new vector to obtain the common features of the class.

[0012] In one possible implementation, the learning feature extraction module extracts error pattern features from the grading data as follows: it extracts all error patterns of each student from the grading data, and extracts the most frequent error patterns from all error patterns of each student to obtain the error type features of each student; then it retrieves the encoding of each item of the error type features of each student and integrates them into a vector form to obtain the error type encoding feature vector of each student.

[0013] In one possible implementation, the learning feature extraction module extracts ability features from the graded data by: calculating the scores of each student in each ability dimension of the teaching syllabus based on the content of the graded data, and integrating them into a vector form to obtain the ability dimension feature vector of each student.

[0014] The diagnostic device, equipped with the aforementioned structure and functions, extracts knowledge mastery features from the graded data to obtain each student's accuracy rate for questions corresponding to each knowledge point, the standard deviation of each student's accuracy rate, and the accuracy rate of each student in analyzing comprehensive questions across multiple knowledge points. Subsequently, it extracts error pattern features from the graded data to obtain each student's error type encoding feature vector. This error root cause analysis allows teachers to improve the targeting of their teaching, avoiding blindly cramming. Finally, it extracts ability features from the graded data to obtain each student's ability dimension feature vector, improving the accuracy of identifying common problems and effectively pinpointing the core weaknesses of students in the class. Simultaneously, by calculating the Hadamard product of each student's error type encoding feature vector and weight vector, it can reduce numerical values, avoid large number operations, and also reduce differences between values, preventing numerical calculation failures due to excessive jumps. Then, it statistically analyzes the mode of each component of the error vector and ability dimension feature vector, integrating the modes of all components of the error vector and ability dimension feature vector into a new vector to obtain the common features of the class. This mode collection method not only reduces the amount of computation but also obtains class commonalities, improving the accuracy of identifying common problems.

[0015] In one possible implementation, the hierarchical analysis device includes: The subgroup partitioning module is electrically connected to the problem diagnosis module. It is used to expand the dimensions of the common features of the class to obtain the universal center vector, and to use spectral clustering to cluster the students in the class according to the learning features of the class, with the universal center vector as the center, so as to divide all students in the class into multiple subgroups. The large language module, electrically connected to the subgroup division module, is used to extract the common weaknesses of students in each subgroup from the grading data, and generate corresponding suggestions for improvement, appropriate resources, and teaching strategies based on the common weaknesses of students in each subgroup.

[0016] The hierarchical analysis device with the above structure and functions expands the dimensions of common features of classes to obtain the universal center vector, and uses spectral clustering to divide subgroups, breaking through the score segment hierarchies. The multi-dimensional subgroup division improves the adaptability of the hierarchical strategy by 5, meets the needs of different students, and makes the hierarchical teaching more refined. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a class learning analysis system based on an intelligent grading system disclosed in an embodiment of this application. Detailed Implementation

[0018] First, those skilled in the art should understand that these embodiments are merely used to explain the technical principles of the embodiments of this application and are not intended to limit the scope of protection of the embodiments of this application. Those skilled in the art can make adjustments as needed to adapt to specific application scenarios.

[0019] In the description of the embodiments of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "electrical connection" and "establishing an electrical connection relationship" should be interpreted broadly, that is, it should be understood that two or more parties have an electrical relationship, which can be achieved through a wire, a radio connection, or a combination of both; it can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of this application based on the specific circumstances.

[0020] The present application will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0021] See Figure 1 This application discloses a class learning analysis system based on an intelligent grading system. Figure 1 This is a schematic diagram of the structure of the class learning analysis system of the department. The class learning analysis system includes an intelligent grading system, a data aggregation device, a diagnostic device, and a hierarchical analysis device. The intelligent grading system has project grading and recording functions and is equipped with an output terminal. The data aggregation device is electrically connected to the output terminal of the intelligent grading system, the diagnostic device is electrically connected to the data aggregation device, and the hierarchical analysis device is electrically connected to the diagnostic device.

[0022] See Figure 1 In this class learning analysis system, the data aggregation device is used to collect the output data of the intelligent grading system in batches and integrate the output data to obtain the grading data of all students in the class. For example... Figure 1 As shown, in this embodiment, the data aggregation device includes a batch data acquisition module and a data preprocessing module. The batch data acquisition module is electrically connected to the output of the intelligent batch system, and the data preprocessing module is electrically connected to both the batch data acquisition module and the diagnostic device.

[0023] In the data aggregation device, the grading data acquisition module is used to collect and output data in batches. In this embodiment, the output data includes question data, student answer data, and basic data containing the number of students, subject, class number, and grading time. In this embodiment, the basic data consists of class number, subject, number of students, and grading time (e.g., "2023-11-05 Math Weekly Test"); the question data includes the knowledge point tag, question type, full score, class accuracy rate, and error type distribution for each question (e.g., "Formula error rate 40%, calculation error rate 30%"); the student answer data includes each student's answer result (correct / incorrect), score, error type, step-by-step score (e.g., "Physics question formula score, step score, result score"), and answering time.

[0024] In the data aggregation device, the data preprocessing module sequentially performs invalid data filtering, dimension alignment, and aggregation calculation on the output data to obtain the graded data of all students in the class. This embodiment filters invalid data (such as absent students or zero-score records due to system errors) through outlier handling; dimension alignment unifies the knowledge point labels for different grading tasks (such as merging "Pythagorean theorem" and "Pythagorean theorem" into the same knowledge point); and aggregation calculation generates class-level statistical indicators (such as the average accuracy rate of knowledge points and the percentage of error types).

[0025] See Figure 1 In this class learning analysis system, the diagnostic device is used to extract knowledge mastery features, error pattern features, and ability features from the graded data to obtain class learning characteristics. Then, these characteristics are used to assess the class's core weaknesses based on a weighted scoring method to obtain common class characteristics. Class learning characteristics include each student's accuracy rate for questions corresponding to each knowledge point, the standard deviation of each student's accuracy rate, each student's accuracy rate in analyzing comprehensive questions across different knowledge points, each student's error type encoding feature vector, and each student's ability dimension feature vector. For example... Figure 1 As shown, in this embodiment, the diagnostic device includes a learning feature extraction module and a problem diagnosis module. The learning feature extraction module communicates with the data preprocessing module, and the problem diagnosis module is electrically connected to both the learning feature extraction module and the hierarchical analysis device.

[0026] In the diagnostic device, the learning feature extraction module is used to extract knowledge mastery features from the graded data to obtain the accuracy rate of each student for each knowledge point, the standard deviation of each student's accuracy rate, and the accuracy rate of each student in analyzing comprehensive questions across knowledge points. Subsequently, error pattern features are extracted from the graded data to obtain the error type encoding feature vector for each student. Finally, ability features are extracted from the graded data to obtain the ability dimension feature vector for each student.

[0027] In this embodiment, the learning feature extraction module extracts error pattern features from the graded data as follows: it extracts all error patterns of each student from the graded data, and then extracts the most frequent error patterns from all error patterns of each student to obtain the error type features of each student. In this embodiment, each error pattern corresponds to a code, and the top 60% of error patterns are extracted. Then, the code of each item of the error type features of each student is retrieved and integrated into a vector form to obtain the error type code feature vector of each student.

[0028] In this embodiment, the learning feature extraction module extracts ability features from the graded data by: calculating the scores of each student in each ability dimension of the teaching syllabus based on the content of the graded data, and integrating them into a vector form to obtain the ability dimension feature vector of each student.

[0029] In the diagnostic device, the problem diagnosis module calculates the Hadamard product of each student's error type encoding feature vector and weight vector to obtain each student's error vector. Then, it calculates the mode of each component of the error vector and the ability dimension feature vector, and integrates the modes of all components of the error vector and the ability dimension feature vector into a new vector to obtain the common features of the class. In this embodiment, the weight vector is set adaptively based on the importance of knowledge points, mastery level, and error concentration. During the generation process, regarding the importance of knowledge points, the core knowledge points are given higher weights than the extended knowledge points; regarding mastery level, the lower the accuracy rate, the higher the weight; and regarding error concentration, the more uniform the error type, the higher the weight.

[0030] See Figure 1 In this class learning analysis system, the stratified analysis device is used to divide all students in the class into multiple subgroups based on common characteristics of the class, and to generate corresponding suggestions for improvement, appropriate resources, and teaching strategies for the common weaknesses of students in each subgroup. For example... Figure 1 As shown, in this embodiment, the hierarchical analysis device includes a subgroup partitioning module and a large language module, wherein the subgroup partitioning module is electrically connected to the problem diagnosis module, and the large language module is electrically connected to the subgroup partitioning module.

[0031] In the hierarchical analysis device, the subgrouping module is used to expand the dimensions of common features of the class to obtain a universal central vector. Then, using spectral clustering, the students in the class are clustered according to the learning characteristics of the class, centered on the universal central vector, to divide all students in the class into multiple subgroups. In this embodiment, the class is divided into 6 subgroups. The spectral clustering method is existing technology and will not be elaborated here. Those skilled in the art can refer to relevant books on data analysis.

[0032] In the hierarchical analysis device, the large language module is used to extract common weaknesses among students in each subgroup from the grading data. Based on these common weaknesses, it generates corresponding suggestions for improvement, appropriate resources, and teaching strategies. The extraction method for common weaknesses involves combining the teaching syllabus with error characteristics to diagnose the possible causes of weaknesses from the knowledge level, the teaching level, and the student level.

[0033] The class learning analysis system based on an intelligent grading system disclosed in this embodiment, building upon existing intelligent grading systems, incorporates a data aggregation device, a diagnostic device, and a hierarchical analysis device. The data aggregation device obtains the grading data of all students in the class, while the diagnostic device extracts knowledge mastery features, error pattern features, and ability features from the grading data to obtain class learning characteristics. These characteristics are then used to assess the class's core weaknesses based on a weighted scoring method to obtain common class characteristics. Because this process utilizes knowledge mastery features, error pattern features, and ability characteristics, it not only focuses on the accuracy of questions but also reflects the assessment of error types and knowledge-related abilities, ensuring educators can promptly grasp teaching effectiveness and prepare lessons accordingly. Simultaneously, the hierarchical analysis device divides all students in the class into multiple subgroups based on these common characteristics and generates corresponding suggestions for improvement, appropriate resources, and teaching strategies for each subgroup's common weaknesses. This process categorizes students based on common class characteristics constructed from knowledge mastery features, error pattern features, and ability characteristics, rather than simple score ranges, effectively avoiding a one-size-fits-all approach in teaching decisions. Furthermore, after subgrouping, suggestions for improvement, appropriate resources, and teaching strategies are generated based on the common weaknesses of students in each subgroup, effectively realizing the educational loop and improving the educational process, thereby meeting the needs of different subgroups.

[0034] In the description of the embodiments of this application, it should be noted that the terms "inner" and "outer" and other terms indicating direction or positional relationship are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and does not indicate or imply that the device or component must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this application.

[0035] In the description of this application, the references to terms such as "an embodiment," "some embodiments," "in this embodiment," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0036] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A class learning analysis system based on an intelligent grading system, comprising an intelligent grading system with project-based grading and recording functions, characterized in that, Also includes: A data aggregation device is electrically connected to the output end of the intelligent grading system, and is used to collect the output data of the intelligent grading system in batches and integrate the output data to obtain the grading data of all students in the class. The diagnostic device, electrically connected to the data aggregation device, is used to extract knowledge mastery features, error pattern features, and ability features from the graded data to obtain class learning features. Then, the class learning features are used to evaluate the class's core weaknesses based on a weighted scoring method to obtain common features of the class. The hierarchical analysis device, electrically connected to the diagnostic device, is used to divide all students in the class into multiple subgroups based on the common characteristics of the class, and to generate corresponding suggestions for improvement, appropriate resources, and teaching strategies for the common weaknesses of students in each subgroup.

2. The class learning analysis system based on the intelligent grading system according to claim 1, characterized in that, The data aggregation device includes: The grading data acquisition module is electrically connected to the output terminal of the intelligent grading system and is used to collect the output data in batches. The output data includes question data, student answer data, and basic data containing the number of students, subject, class number, and grading time. The data preprocessing module is electrically connected to both the graded data acquisition module and the diagnostic device. It is used to sequentially perform invalid data filtering, dimension alignment, and aggregation calculation on the output data to obtain the graded data of all students in the class.

3. The class learning analysis system based on an intelligent grading system according to claim 1 or 2, characterized in that, The class learning characteristics include each student's accuracy rate for questions corresponding to each knowledge point, the standard deviation of each student's accuracy rate, the accuracy rate of each student in analyzing comprehensive questions across knowledge points, the error type encoding feature vector of each student, and the ability dimension feature vector of each student.

4. The class learning analysis system based on the intelligent grading system according to claim 3, characterized in that, The diagnostic device includes: The learning feature extraction module communicates with the data preprocessing module to extract knowledge mastery features from the graded data, thereby obtaining the accuracy rate of each student for each knowledge point, the standard deviation of each student's accuracy rate, and the accuracy rate of each student in analyzing comprehensive questions across knowledge points. Subsequently, error pattern features are extracted from the graded data to obtain the error type encoding feature vector for each student. Finally, ability features are extracted from the graded data to obtain the ability dimension feature vector for each student. The problem diagnosis module is electrically connected to both the learning feature extraction module and the hierarchical analysis device. It is used to calculate the Hadamard product of the error type encoding feature vector and the weight vector of each student to obtain the error vector of each student. Then, it calculates the mode of each component of the error vector and the ability dimension feature vector, and integrates the modes of all components of the error vector and the ability dimension feature vector into a new vector to obtain the common features of the class.

5. The class learning analysis system based on the intelligent grading system according to claim 4, characterized in that, The learning feature extraction module extracts error pattern features from the graded data in the following way: it extracts all error patterns of each student from the graded data, and extracts the most frequent error patterns from all error patterns of each student to obtain the error type features of each student. Then, the code of each item of the error type feature of each student is retrieved and integrated into a vector form to obtain the error type code feature vector of each student.

6. The class learning analysis system based on the intelligent grading system according to claim 4, characterized in that, The learning feature extraction module extracts ability features from the graded data by: calculating the scores of each student in each ability dimension of the teaching syllabus based on the content of the graded data, and integrating them into a vector form to obtain the ability dimension feature vector of each student.

7. The class learning analysis system based on an intelligent grading system according to any one of claims 4-6, characterized in that, The hierarchical analysis device includes: The subgroup partitioning module is electrically connected to the problem diagnosis module. It is used to expand the dimensions of the common features of the class to obtain the universal center vector, and to use spectral clustering to cluster the students in the class according to the learning features of the class, with the universal center vector as the center, so as to divide all students in the class into multiple subgroups. The large language module, electrically connected to the subgroup division module, is used to extract the common weaknesses of students in each subgroup from the grading data, and generate corresponding suggestions for improvement, appropriate resources, and teaching strategies based on the common weaknesses of students in each subgroup.

Images