Automatic learning condition analysis system based on AI intelligence

Abstract

The invention discloses an AI intelligence-based learning condition automatic analysis system, which relates to the technical field of learning condition analysis and comprises a learning condition data preprocessing module, a learning condition factor generation module, a learning condition factor clustering module, a learning condition factor comparison and analysis module, a learning condition factor association module, a learning condition diagnosis and evaluation module and an eye movement tracking module. According to the invention, through the arrangement of the eye movement tracking module, the teacher end can accurately master points which are difficult to understand when students read, so that subsequent targeted tutoring is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of learning situation analysis, and more specifically to an automatic learning situation analysis system based on AI intelligence. Background Art

[0002] The automatic learning situation analysis system uses big data and artificial intelligence technology to collect students' learning behavior data in real time from multiple channels such as learning management systems (LMS), online course platforms, and smart classroom equipment. It uses advanced machine learning and data mining algorithms for analysis and displays the analysis results in the form of charts. It provides educators with insights into students' learning status and can provide students with personalized learning resource recommendations based on the analysis results.

[0003] When studying, reading articles, questions, etc. is the first step in learning. Then, when reading, students' concentration and reading efficiency become a factor that affects subsequent learning. Existing analysis systems usually analyze students' learning process based on some outcome data of students. However, reading accounts for a larger proportion in learning. By analyzing students' reading process, it is more conducive to mastering their comprehension ability.

[0004] Therefore, it is necessary to propose an automatic learning situation analysis system based on AI intelligence to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to solve the problems existing in the background technology.

[0006] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0007] An AI-based automatic learning situation analysis system, including a student terminal and a teacher terminal connected to a background processing module, where both students and teachers can log in with a unique student ID and a self-set password;

[0008] It also includes a learning data collection module that integrates multiple data sources, including but not limited to student homework submission records, classroom interactive video viewing behavior, online test scores, questionnaire feedback, reading concentration, and reading efficiency;

[0009] The learning data preprocessing module cleans the data through density-based outlier detection to remove noise and outliers; it converts the data through normalization to convert data in different formats into a unified format and eliminate the impact of dimension;

[0010] The learning factor generation module uses feature extraction methods to extract key features from the data processed by the learning data preprocessing module as learning factors, and reduces and combines the features to generate more representative learning factors.

[0011] The learning factor clustering module uses clustering algorithms to perform cluster analysis on learning factors and divides students into different learning groups based on different learning data sources;

[0012] The learning situation factor comparison analysis module compares and analyzes the learning situation factors of the same student group at different time points, and compares and analyzes the learning situation factors of different student groups at the same time point;

[0013] The learning factor association module conducts association analysis on learning factors and discovers strong association rules between factors;

[0014] The learning situation diagnosis and assessment module accurately diagnoses students' learning status and evaluates their mastery of knowledge points by comparing their actual mastery and predicted mastery. It predicts students' future learning performance and development potential by comprehensively analyzing the changing trends and correlations of learning situation factors, and finally generates a personalized learning situation diagnosis report.

[0015] Furthermore, the learning data collection module also includes an eye tracking module, which is used to sequentially generate four guiding symbols at the four corners of the screen viewed by the student, so as to guide the student's eyeballs to scan the four corners in sequence, and then record the effective eyeball scanning path of the student's learning this time;

[0016] When students are reading, the eye tracking module monitors the students' eyes to generate an eye reading trajectory, wherein when the students' eyes leave the effective eye scanning path, the eye reading trajectory is disconnected within the time period until the eye tracking module monitors that the students' eyes fall back into the effective eye scanning path.

[0017] Furthermore, the eye movement reading trajectory is arranged in multiple rows, and the time required to form each row of eye movement reading trajectory is marked. When the eye movement reading trajectory is disconnected within the time period and falls into the effective scanning path of the eye again, the trajectory continues to be generated from the eye movement reading trajectory of the disconnected row, and the required time includes the time from disconnection to falling into the effective scanning path of the eye again. Finally, the total reading time and the number of disconnections are generated as learning situation factors.

[0018] Furthermore, when analyzing reading concentration and reading efficiency, the learning situation factor comparison and analysis module classifies and ranks student groups according to the number of disconnections in the eye movement reading trajectory and the total reading time generated ultimately, and ranks them by the number of disconnections and reading time respectively, and transmits the eye movement reading trajectories of students with lower rankings to the learning situation factor comparison and analysis module through self-set disconnection number standards and reading time standards.

[0019] Furthermore, when analyzing the reading concentration, the learning situation factor comparison and analysis module performs correlation analysis based on the number of disconnections and the learning situation factors obtained from other data sources;

[0020] When analyzing reading efficiency, the learning situation factor comparison and analysis module calculates the average reading speed of the student based on the dynamic eye movement reading trajectory, and then analyzes the part of each row of the dynamic eye movement reading trajectory where the speed is lower than the average as a mark;

[0021] The learning situation factor association module performs association analysis based on the marked and below-average parts, including but not limited to rare characters and beyond the cognitive range. Finally, the learning situation diagnosis and evaluation module generates a reading efficiency diagnosis report and transmits the results to the teacher.

[0022] A row of eye movement reading trajectories is set to correspond to a row of reading content. By analyzing the reading speed that is lower than the average reading speed, the position of the students' reading content is located, so that the teacher can accurately analyze the factors affecting students' reading.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] The present invention, through the setting of the eye tracking module, can enable the teacher to accurately grasp the points that are difficult for students to understand when reading, thereby facilitating subsequent targeted tutoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a simple diagram of the eye movement reading trajectory. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] Combination Figure 1 As shown, an AI-based automatic learning situation analysis system includes a student end and a teacher end that are connected to the background processing module. Both students and teachers can log in with a unique student ID and a self-set password. The teacher end is mainly responsible for timely updating the online learning content according to the students' offline learning situation, and finally performs manual analysis based on the data diagnosed by the learning situation diagnosis and evaluation module to analyze the main influencing factors and weaknesses of the students.

[0028] In addition, the automatic analysis system also includes a learning data collection module, which integrates multiple data sources, including but not limited to student homework submission records, classroom interactive video viewing behavior, online test scores, questionnaire feedback, reading concentration and reading efficiency, etc., to comprehensively analyze students' learning situation;

[0029] The learning data preprocessing module cleans the data through density-based outlier detection to remove noise and outliers; converts the data through normalization to convert data in different formats into a unified format and eliminate the impact of dimension; reduces the dimension of the data through principal component analysis to reduce data redundancy, improve the efficiency of subsequent processing, and retain key information;

[0030] The learning factor generation module uses feature extraction methods to extract key features from the data processed by the learning data preprocessing module as learning factors, and reduces and combines the features to generate more representative learning factors.

[0031] The learning factor clustering module uses clustering algorithms to perform cluster analysis on learning factors and divides students into different learning groups based on different learning data sources;

[0032] The choice of clustering algorithm can be optimized according to the characteristics of the data and the clustering goal, such as evaluating the clustering effect through the silhouette coefficient. Through cluster analysis, the commonalities and differences of different groups in learning factors can be found, providing strong support for teaching improvement.

[0033] The learning situation factor comparison analysis module compares and analyzes the learning situation factors of the same student group at different time points, and compares and analyzes the learning situation factors of different student groups at the same time point. That is, it analyzes students of different groups based on different learning situation factors, and analyzes the standards of the learning situation factors, so as to facilitate the subsequent targeted auxiliary corrections by teachers;

[0034] The learning factor association module conducts association analysis on learning factors and discovers strong association rules between factors. Through association analysis, the causal relationship and mutual influence between learning factors can be revealed;

[0035] The learning situation diagnosis and assessment module accurately diagnoses students' learning status and evaluates their mastery of knowledge points by comparing their actual mastery and predicted mastery. It predicts students' future learning performance and development potential by comprehensively analyzing the changing trends and correlations of learning situation factors. Finally, it generates a personalized learning situation diagnosis report, which can be viewed on the teacher client. For students with poor learning performance, a more conspicuous way can be used to remind teachers to review it in time.

[0036] Specifically, for example, in terms of students' reading concentration and reading efficiency, this is the first step of students' learning. For example, when reading an essay or a question, when students have enough attention to read, the subsequent recitation or analysis can be more in-depth. Therefore, the learning data collection module is also provided with an eye tracking module, which is used to sequentially generate four guiding symbols at the four corners of the screen viewed by the students, so as to guide the students' eyes to scan the four corners in turn, and then record the effective eye scanning path of the students' learning this time. The guidance can be provided before the user reads each time, or according to the device used by the user, the inclination angle of the device can be analyzed according to the gyroscope inside the device, so as to guide and record the effective eye scanning path of the students' learning this time when the inclination angle changes;

[0037] When students are reading, the eye tracking module monitors the students' eyes to generate an eye reading trajectory. When the students' eyes leave the effective eye scanning path, the eye reading trajectory is disconnected within the time period until the eye tracking module monitors that the students' eyes fall back into the effective eye scanning path. In this solution, the reading time and the number of disconnections are recorded to facilitate subsequent analysis.

[0038] Specifically, the eye movement reading trajectory is arranged in multiple rows, and the time required to form each row of eye movement reading trajectory is marked. When the eye movement reading trajectory is disconnected within the time period and falls into the effective scanning path of the eye again, the trajectory continues to be generated from the eye movement reading trajectory of the disconnected row, and the required time includes the time from disconnection to falling into the effective scanning path of the eye again. Finally, the total reading time and the number of disconnections are generated as learning situation factors.

[0039] Specifically, when analyzing reading concentration and reading efficiency, the learning situation factor comparison and analysis module classifies and ranks student groups according to the number of disconnections in the eye movement reading trajectory and the total reading time finally generated, and ranks them according to the number of disconnections and reading time respectively. Through the self-set number of disconnections and reading time standards, the eye movement reading trajectories of the bottom-ranked students are transmitted to the learning situation factor comparison and analysis module, wherein the number of disconnections indicates that the students are not focused enough when reading, and the long total reading time indicates that the students spend a long time reading a certain line, which can also indicate whether the students encounter difficult-to-understand knowledge when reading this line of the article. By recording these two aspects and cooperating with the teacher's targeted guidance, it is possible to effectively discover the students' problems and improve their abilities in a targeted manner.

[0040] Specifically, when analyzing reading concentration, the learning situation factor comparison and analysis module performs correlation analysis based on the number of disconnections and the learning situation factors obtained from other data sources;

[0041] The learning factor comparison and analysis module calculates the average reading speed of the student according to the dynamic eye movement reading trajectory when analyzing reading efficiency. The calculation can be performed according to the previous reading speed, and then the part of each row of dynamic eye movement reading trajectory with a speed lower than the average is analyzed as a mark;

[0042] The learning situation factor association module performs association analysis based on the marked and lower-than-average parts, including but not limited to rare characters and beyond-cognitive range. Finally, a reading efficiency diagnosis report is generated through the learning situation diagnosis and evaluation module, and the results are transmitted to the teacher. The teacher needs to combine the report content and the corresponding reading part to watch the student's dynamic eye movement reading trajectory, so as to intuitively see the position where the eye movement reading trajectory is slower, analyze the content in the reading part according to the position, and provide targeted tutoring;

[0043] like Figure 1 As shown, the node on the right end of the horizontal row represents the deadline, and the broken horizontal line represents the number of disconnections. A row of eye movement reading trajectories is set to correspond to a row of reading content. By analyzing the reading speed that is lower than the average reading speed, the position of the students' reading content is located, so that the teacher can accurately analyze the factors affecting students' reading.

[0044] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. The patent protection scope of the present invention shall be based on the claims. All equivalent structural changes made using the contents of the description and drawings of the present invention should also be included in the protection scope of the present invention.

Claims

1. An AI-based automatic learning situation analysis system, characterized by: Including the student side and teacher side that communicate with the background processing module. Both students and teachers can log in with their unique student ID and self-set password; It also includes a learning data collection module that integrates multiple data sources, including but not limited to student homework submission records, classroom interactive video viewing behavior, online test scores, questionnaire feedback, reading concentration, and reading efficiency; The learning data preprocessing module cleans the data through density-based outlier detection to remove noise and outliers; Transform data by normalization, convert data in different formats into a unified format, and eliminate the impact of dimensions; The learning factor generation module uses feature extraction methods to extract key features from the data processed by the learning data preprocessing module as learning factors, and reduces and combines the features to generate more representative learning factors. The learning factor clustering module uses clustering algorithms to perform cluster analysis on learning factors and divides students into different learning groups based on different learning data sources; The learning situation factor comparison analysis module compares and analyzes the learning situation factors of the same student group at different time points, and compares and analyzes the learning situation factors of different student groups at the same time point; The learning factor association module conducts association analysis on learning factors and discovers strong association rules between factors; The learning situation diagnosis and assessment module accurately diagnoses students' learning status and evaluates their mastery of knowledge points by comparing their actual mastery and predicted mastery. It predicts students' future learning performance and development potential by comprehensively analyzing the changing trends and correlations of learning situation factors, and finally generates a personalized learning situation diagnosis report.

2. According to claim 1, the AI-based automatic learning situation analysis system is characterized by: The learning data collection module also includes an eye tracking module, which is used to sequentially generate four guiding symbols at the four corners of the screen viewed by the student, so as to guide the student's eyeballs to scan the four corners in turn, and then record the effective eyeball scanning path of the student's learning this time; When students are reading, the eye tracking module monitors the students' eyes to generate an eye reading trajectory, wherein when the students' eyes leave the effective eye scanning path, the eye reading trajectory is disconnected within the time period until the eye tracking module monitors that the students' eyes fall back into the effective eye scanning path.

3. The AI-based automatic learning situation analysis system according to claim 2 is characterized by: The eye movement reading trajectory is arranged in multiple rows, and the time required to form each row of eye movement reading trajectory is marked. When the eye movement reading trajectory is disconnected within the time period and falls into the effective scanning path of the eye again, the trajectory continues to be generated from the eye movement reading trajectory of the disconnected row, and the required time includes the time from disconnection to falling into the effective scanning path of the eye again. Finally, the total reading time and the number of disconnections are generated as learning situation factors.

4. The AI-based automatic learning situation analysis system according to claim 3 is characterized by: When analyzing reading concentration and reading efficiency, the learning situation factor comparison and analysis module classifies and ranks student groups according to the number of disconnections in the eye movement reading trajectory and the total reading time generated, and ranks them according to the number of disconnections and reading time respectively. The eye movement reading trajectories of students with low rankings are transmitted to the learning situation factor comparison and analysis module based on the self-set number of disconnections and reading time standards.

5. The AI-based automatic learning situation analysis system according to claim 4 is characterized by: The learning situation factor comparison and analysis module performs correlation analysis based on the number of disconnections and the learning situation factors obtained from other data sources when analyzing reading concentration; When analyzing reading efficiency, the learning situation factor comparison and analysis module calculates the average reading speed of the student based on the dynamic eye movement reading trajectory, and then analyzes the part of each row of the dynamic eye movement reading trajectory where the speed is lower than the average as a mark; The learning situation factor association module performs association analysis based on the marked and below-average parts, including but not limited to rare characters and beyond the cognitive range. Finally, the learning situation diagnosis and evaluation module generates a reading efficiency diagnosis report and transmits the results to the teacher. A row of eye movement reading trajectories is set to correspond to a row of reading content. By analyzing the reading speed that is lower than the average reading speed, the position of the students' reading content is located, so that the teacher can accurately analyze the factors affecting students' reading.

Images