System and method for analyzing and predicting scores based on learning behaviors of students

By collecting and analyzing students and class learning behavior data in the grade prediction system, and using a fusion algorithm model to predict grades, the problem of large prediction errors in the existing technology is solved, and the accuracy and monitoring effect of grade prediction are improved.

CN119026950APending Publication Date: 2024-11-26CHONGQING NORMAL UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411110838.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing grade prediction system has errors in the detailed collection of students and class learning behaviors, and it is difficult to effectively monitor the changes in grades, resulting in large prediction errors.

Method used

A system based on student learning behavior analysis and prediction results was designed, including learning behavior collection module, induction and sorting module, learning behavior summary analysis module and student learning behavior prediction module to collect offline and online learning behavior data and build learning behavior data. and achievement predictions are made through fusion algorithm models such as polynomial Naive Bayes, AdaBoost, Gradientboosting, and Logistic Regression.

Benefits of technology

By collecting and analyzing students and class learning behavior data in detail, the error of grade prediction is reduced and the accuracy and effectiveness of grade change monitoring is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026950A_ABST
    Figure CN119026950A_ABST
Patent Text Reader

Abstract

The invention discloses a system and a method for analyzing and predicting scores based on student learning behaviors. The system comprises a learning behavior acquisition module used for acquiring offline class conditions, library clock-in conditions and online class clock-in conditions; the induction and sorting module is used for carrying out induction and classification on offline classroom behavior data, learning behavior data of a library and learning conditions of online lessons, and sorting corresponding learning behavior data models according to each student and each class; and the learning behavior summarizing and analyzing module is used for summarizing and analyzing the learning behavior data of each student and each class. According to the invention, the score table is constructed according to the usual score, the score of each chapter, the score in the period and the score at the end of the period of each student, the score change condition of the student can be analyzed through the learning behavior of the student, and the class score table is constructed according to the score of each year of each class. And the score change condition of each class is analyzed through the score linear change of each class.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of student performance prediction, and in particular to a system and method for predicting performance based on student learning behavior analysis. Background Art

[0002] In order to understand and predict each student's learning situation, a performance prediction system can be used to predict students' performance through analysis of their learning behavior.

[0003] In the prior art, a method for predicting student rankings based on campus data is disclosed in the publication number CN105894119A. The method collects data of all students in school, including grade data and behavior data, cleans the student data, and normalizes non-time data items. The behavior feature vector of each student is extracted from the processed data. The behavior feature includes grade feature, effort feature, and life rule feature. The behavior feature vector is then reduced in dimension. Each student uses its reduced-dimensional behavior feature vector to subtract the behavior feature vector of each other student to obtain a difference feature vector, which is input into a classifier to obtain a corresponding label value. The label values ​​are summed to obtain the student's score. The scores of all students are sorted to obtain the ranking prediction value of each student. The invention analyzes the student's campus data, uses data to describe the student's learning habits and behavior characteristics, and predicts the student's ranking as a reference for student education. The performance prediction system has the following problems:

[0004] Generally, robots are used to analyze students' academic performance and behavior, which is not convenient for collecting detailed learning behaviors of each student and each class. Then, predicting students' grades is prone to large grade prediction errors and is not convenient for effectively monitoring students' grade changes. To this end, we propose a system and method for predicting grades based on student learning behavior analysis. Summary of the invention

[0005] The purpose of the present invention is to provide a system and method for predicting grades based on student learning behavior analysis, which solves the existing problems.

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

[0007] A system for predicting grades based on student learning behavior analysis, comprising:

[0008] The learning behavior collection module is used to collect offline class attendance, library clock-in status, and online class clock-in status;

[0009] The summarization and sorting module is used to summarize and classify offline classroom behavior data, library learning behavior data, and online course learning conditions, and sort out corresponding learning behavior data models for each student and each class;

[0010] The learning behavior summary analysis module is used to summarize and analyze the learning behavior data of each student and each class;

[0011] The student learning behavior prediction module is used to construct basic learning feature data from the students' mid-term and final exam scores, and input the basic learning feature data into the trained fusion algorithm model to obtain the student score prediction results;

[0012] The storage device is used to collect the learning behavior data of each student and each class, and store the learning performance of each student and each class in each subject at each stage;

[0013] The data transmission module is used to transmit students' learning behavior data and academic performance to the local host.

[0014] Preferred offline classes:

[0015] Create a timetable for each student, record the attendance of each class, and then create a separate attendance table for each student based on the attendance of each student per quarter, half year, and full year. Then summarize the attendance of each class and collect the students' offline grades per quarter, half year, and full year.

[0016] Preferred, offline scores:

[0017] Collect each student's learning behavior data on chapter test scores, midterm test scores, and final exam scores.

[0018] Preferably, the library's card-checking situation:

[0019] Record students' clock-in status at the library, and record the books borrowed each time from the library, as well as the time spent in the library.

[0020] Preferably, the online course check-in situation:

[0021] Collect each student's online class check-in status, and collect the test results of each class, each chapter, mid-term, and final exams, as well as the corresponding learning behavior data, and collect the teacher's scores for each student.

[0022] Preferably, the learning behavior summary analysis module includes:

[0023] The individual student analysis module is used to analyze each student's grades in each subject, class attendance, and further analyze the learning behavior of individual students;

[0024] The class performance analysis module is used to analyze the performance of each class in each subject, the class attendance, and then analyze the learning behavior of each class;

[0025] Linear analysis module: construct a grade table with each student's regular grades in each subject, grades in each chapter, mid-term grades, and final grades. Analyze the changes in students' grades through their learning behaviors, and construct a class grade table with each class's annual grades. Then analyze the changes in each class's grades through the linear changes in each class's grades.

[0026] Preferably, the fusion algorithm model includes a primary classification model constructed based on three classification models: multinomial naive Bayes, AdaBoost and Gradientboosting, and a secondary classification model constructed based on a logistic regression model.

[0027] Preferably, the fusion algorithm model includes:

[0028] The linear correlation between the preprocessed basic learning feature data and the target feature variable is calculated by the Pierce correlation coefficient, and the basic learning feature data that is not linearly correlated with the target feature variable is deleted to obtain a processed student feature data set;

[0029] Divide the processed student feature data set into a training set and a test set;

[0030] The training set is used to train the primary classification model and the secondary classification model in the fusion algorithm model to obtain a trained fusion algorithm model.

[0031] A systematic prediction method for predicting grades based on student learning behavior analysis includes the following steps:

[0032] Step S1, the learning behavior collection module is used to collect the attendance of each quarter, each half year, and the whole year, establish a separate attendance table for each student, and then summarize the attendance of each class, as well as each student's offline grades collected every quarter, each half year, and the whole year, and collect the students' clock-in status at the library, as well as the books borrowed and the time in the library, and collect the online class clock-in status of each student, and collect the test results of each class, each chapter, mid-term, and final, as well as the corresponding learning behavior data, and collect the teacher's score for each student;

[0033] Step S2, transmitting the learning behavior data collected by the learning behavior collection module to the summarization and sorting module, and using the summarization and sorting module to summarize and classify the offline classroom behavior data, the library learning behavior data, and the online course learning situation, and sort out the corresponding learning behavior data model according to each student and each class, and then transmit the summarized behavior data to the learning behavior summary and analysis module;

[0034] Step S3, the learning behavior summary analysis module analyzes the grades of each subject and the class situation of each student and each class through the individual student analysis module and the class score analysis module, and then analyzes the learning behavior of the students and the class, and constructs a score table with the regular score, chapter score, mid-term score, and final score of each student in each subject, and can analyze the change of students' grades through the students' learning behavior, and construct a class score table with the annual grades of each class, and then analyze the change of grades of each class through the linear change of the grades of each class;

[0035] Step S4: The learning behavior data of students' mid-term and final exam scores are used to construct basic learning feature data through the student learning behavior prediction module, and the basic learning feature data is input into the trained fusion algorithm model to obtain the student score prediction results.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] The present invention uses a learning behavior collection module to collect offline class conditions, library check-in conditions, and online class check-in conditions, and uses a summarizing and sorting module to summarize and classify offline classroom behavior data, library learning behavior data, and online class learning conditions, and sorts out corresponding learning behavior data models according to each student and each class, and uses a student learning behavior prediction module to collect the learning behavior data of students' mid-term and final exam scores in various subjects to construct basic learning feature data, and input the basic learning feature data into a trained fusion algorithm model to obtain student performance prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the structure of the present invention;

[0039] Figure 2 It is a structural schematic diagram of the learning behavior summary and analysis module of the present invention. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0041] like Figure 1 and Figure 2 As shown, a system for predicting grades based on student learning behavior analysis includes:

[0042] The learning behavior collection module is used to collect offline class attendance, library clock-in status, and online class clock-in status;

[0043] The summarization and sorting module is used to summarize and classify offline classroom behavior data, library learning behavior data, and online course learning conditions, and sort out corresponding learning behavior data models for each student and each class;

[0044] The learning behavior summary analysis module is used to summarize and analyze the learning behavior data of each student and each class;

[0045] The student learning behavior prediction module is used to construct basic learning feature data from the students' mid-term and final exam scores, and input the basic learning feature data into the trained fusion algorithm model to obtain the student score prediction results;

[0046] The storage device is used to collect the learning behavior data of each student and each class, and store the learning performance of each student and each class in each subject at each stage;

[0047] The data transmission module is used to transmit students' learning behavior data and academic performance to the local host.

[0048] Preferably, the fusion algorithm model includes a primary classification model constructed based on three classification models: multinomial naive Bayes, AdaBoost and Gradientboosting, and a secondary classification model constructed based on a logistic regression model.

[0049] A systematic prediction method for predicting grades based on student learning behavior analysis includes the following steps:

[0050] Step S1, the learning behavior collection module is used to collect the attendance of each quarter, each half year, and the whole year, establish a separate attendance table for each student, and then summarize the attendance of each class, as well as each student's offline grades collected every quarter, each half year, and the whole year, and collect the students' clock-in status at the library, as well as the books borrowed and the time in the library, and collect the online class clock-in status of each student, and collect the test results of each class, each chapter, mid-term, and final, as well as the corresponding learning behavior data, and collect the teacher's score for each student;

[0051] Step S2, transmitting the learning behavior data collected by the learning behavior collection module to the summarization and sorting module, and using the summarization and sorting module to summarize and classify the offline classroom behavior data, the library learning behavior data, and the online course learning situation, and sort out the corresponding learning behavior data model according to each student and each class, and then transmit the summarized behavior data to the learning behavior summary and analysis module;

[0052] Step S3, the learning behavior summary analysis module analyzes the grades of each subject and the class situation of each student and each class through the individual student analysis module and the class score analysis module, and then analyzes the learning behavior of the students and the class, and constructs a score table with the regular score, chapter score, mid-term score, and final score of each student in each subject, and can analyze the change of students' grades through the students' learning behavior, and construct a class score table with the annual grades of each class, and then analyze the change of grades of each class through the linear change of the grades of each class;

[0053] Step S4: The learning behavior data of the students' mid-term and final exam scores are used to construct basic learning feature data through the student learning behavior prediction module, and the basic learning feature data is input into the trained fusion algorithm model to obtain the student score prediction results;

[0054] Step S41, the fusion algorithm model includes:

[0055] The linear correlation between the preprocessed basic learning feature data and the target feature variable is calculated by the Pierce correlation coefficient, and the basic learning feature data that is not linearly correlated with the target feature variable is deleted to obtain a processed student feature data set;

[0056] Divide the processed student feature data set into a training set and a test set;

[0057] The training set is used to train the primary classification model and the secondary classification model in the fusion algorithm model to obtain a trained fusion algorithm model.

[0058] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A system for predicting grades based on student learning behavior analysis, It is characterized in that include: The learning behavior collection module is used to collect offline class attendance, library clock-in status, and online class clock-in status; The summarization and sorting module is used to summarize and classify offline classroom behavior data, library learning behavior data, and online course learning conditions, and sort out corresponding learning behavior data models for each student and each class; The learning behavior summary analysis module is used to summarize and analyze the learning behavior data of each student and each class; The student learning behavior prediction module is used to construct basic learning feature data from the students' mid-term and final exam scores, and input the basic learning feature data into the trained fusion algorithm model to obtain the student score prediction results; The storage device is used to collect the learning behavior data of each student and each class, and store the learning performance of each student and each class in each subject at each stage; The data transmission module is used to transmit students' learning behavior data and academic performance to the local host.

2. A system for predicting grades based on student learning behavior analysis according to claim 1, It is characterized in that Offline class situation: Create a timetable for each student, record the attendance of each class, and then create a separate attendance table for each student based on the attendance of each student per quarter, half year, and full year. Then summarize the attendance of each class and collect the students' offline grades per quarter, half year, and full year.

3. A system for predicting grades based on student learning behavior analysis according to claim 1, It is characterized in that Offline results: Collect each student's learning behavior data on chapter test scores, midterm test scores, and final exam scores.

4. A system for predicting grades based on student learning behavior analysis according to claim 1, It is characterized in that Library card punching situation: Record students' clock-in status at the library, and record the books borrowed each time from the library, as well as the time spent in the library.

5. A system for predicting grades based on student learning behavior analysis according to claim 1, It is characterized in that Online course check-in status: Collect each student's online class check-in status, and collect the test results of each class, each chapter, mid-term, and final exams, as well as the corresponding learning behavior data, and collect the teacher's scores for each student.

6. A system for predicting grades based on student learning behavior analysis according to claim 1, It is characterized in that The learning behavior summary analysis module includes: The individual student analysis module is used to analyze each student's grades in each subject, class attendance, and further analyze the learning behavior of individual students; The class performance analysis module is used to analyze the performance of each class in each subject, the class attendance, and then analyze the learning behavior of each class; Linear analysis module: construct a grade table with each student's regular grades in each subject, grades in each chapter, mid-term grades, and final grades. Analyze the changes in students' grades through their learning behaviors, and construct a class grade table with each class's annual grades. Then analyze the changes in each class's grades through the linear changes in each class's grades.

7. A system for predicting grades based on student learning behavior analysis according to claim 1, It is characterized in that The fusion algorithm model includes a primary classification model constructed based on three classification models: multinomial naive Bayes, AdaBoost and Gradientboosting, and a secondary classification model constructed based on a logistic regression model.

8. A system for predicting grades based on student learning behavior analysis according to claim 7, It is characterized in that The fusion algorithm model includes: The linear correlation between the preprocessed basic learning feature data and the target feature variable is calculated by the Pierce correlation coefficient, and the basic learning feature data that is not linearly correlated with the target feature variable is deleted to obtain a processed student feature data set; Divide the processed student feature data set into a training set and a test set; The training set is used to train the primary classification model and the secondary classification model in the fusion algorithm model to obtain a trained fusion algorithm model.

9. A prediction method for a system for predicting grades based on student learning behavior analysis according to any one of claims 1 to 8, It is characterized in that The following steps are involved: Step S1, the learning behavior collection module is used to collect the attendance of each quarter, each half year, and the whole year, establish a separate attendance table for each student, and then summarize the attendance of each class, as well as each student's offline grades collected every quarter, each half year, and the whole year, and collect the students' clock-in status at the library, as well as the books borrowed and the time in the library, and collect the online class clock-in status of each student, and collect the test results of each class, each chapter, mid-term, and final, as well as the corresponding learning behavior data, and collect the teacher's score for each student; Step S2, transmitting the learning behavior data collected by the learning behavior collection module to the summarization and sorting module, and using the summarization and sorting module to summarize and classify the offline classroom behavior data, the library learning behavior data, and the online course learning situation, and sort out the corresponding learning behavior data model according to each student and each class, and then transmit the summarized behavior data to the learning behavior summary and analysis module; Step S3, the learning behavior summary analysis module analyzes the grades of each subject and the class situation of each student and each class through the individual student analysis module and the class score analysis module, and then analyzes the learning behavior of the students and the class, and constructs a score table with the regular score, chapter score, mid-term score, and final score of each student in each subject, and can analyze the change of students' grades through the students' learning behavior, and construct a class score table with the annual grades of each class, and then analyze the change of grades of each class through the linear change of the grades of each class; Step S4: The learning behavior data of students' mid-term and final exam scores are used to construct basic learning feature data through the student learning behavior prediction module, and the basic learning feature data is input into the trained fusion algorithm model to obtain the student score prediction results.