A method, system, medium, and device for personalized learning path planning based on user behavior.

Abstract

This invention relates to the field of personalized learning path planning, and discloses a method and system for personalized learning path planning based on user behavior. The method includes: after completing a phase of course learning, users engage in learning behaviors such as video viewing, simulated practice, and forum discussions; the system's backend database is used to obtain the state of these behaviors, and the weight scores of different behaviors are analyzed to construct a user behavior state model; based on the user's behavior state and the internal weight scores of the model, the user's learning outcomes are estimated and judged; and a personalized learning path is generated based on the estimated and judged results. This invention analyzes user learning behavior states and establishes personalized learning paths, solving the problem of low matching between previous fixed learning paths and user learning abilities, improving user learning efficiency, and providing a more accurate and suitable learning experience for different users.

Description

Technical Field

[0001] This invention relates to the field of personalized education technology, specifically to a method, system, medium, and device for planning personalized learning paths based on user behavior. Background Technology

[0002] Online education and training platforms offer a wealth of learning resources, covering everything from basic knowledge to advanced skills. Users can find the learning content they need from these resources. However, as users learn, their needs become increasingly diverse and personalized. For example, different users have different requirements regarding the difficulty, format, pace, and duration of the learning content. Traditional course recommendation systems often rely on fixed learning paths or pre-set course combinations. This method struggles to adapt to these diverse needs. Combined with the inherent differences in users' personalities and their varying levels of understanding, different users using traditional fixed learning paths often experience vastly different levels of motivation and efficiency when learning a particular course. This results in existing courses failing to meet users' personalized needs.

[0003] Personalized learning paths, as a popular research direction in education and training, are rapidly changing traditional teaching models and providing users with unique and effective learning experiences. Currently, research in the field of personalized learning paths mainly focuses on user personality characteristics (such as learning styles). This approach focuses on the user's inherent learning characteristics and achieves personalized education by inferring the user's learning preferences. However, it ignores the user's learning behavior and the degree of mastery of course knowledge points. Existing personalized learning systems mainly use ant colony algorithms to improve the matching degree between courses and users. However, ant colony algorithms tend to stagnate in the neighborhood of one or more local optima, resulting in premature convergence. That is, after the search has progressed to a certain extent, all the solutions found by all ants are completely consistent, and it is impossible to continue to search the solution space further. In personalized learning, users' knowledge level, interests, and learning goals are often constantly changing, which leads to a low fit rate of the learning resources recommended by the algorithm or the learning path planned for the learner, resulting in a poor personalized experience for the user. Summary of the Invention

[0004] To address the above problems, this invention proposes a personalized learning path planning method, system, medium, and device based on user behavior. This method collects user learning behaviors generated during the learning process, constructs a user behavior state model, evaluates user learning outcomes, establishes personalized learning paths, improves course matching, enhances user learning efficiency, and increases user learning achievement. Therefore, to achieve the above technical effects, this invention provides the following technical solution: A personalized learning path planning method based on user behavior, comprising the following steps:

[0005] Step S1: Collect user learning behavior data based on the system's backend database. The user behavior data includes video viewing data, simulation practice data, and forum communication data.

[0006] According to the above technical solution, the video viewing data further includes the number of video plays and the total playback duration; the simulation practice data further includes the simulation practice score; and the forum communication data further includes the number of forum posts and comments.

[0007] Step S2: Classify the collected user learning behavior data and construct a user behavior state model based on the classification results. The user behavior state model consists of course completion rate, course mastery rate, and community participation rate.

[0008] According to the above technical solution, classifying the collected user learning behavior data includes: classifying video viewing behavior into course completion status, simulated practice behavior into course mastery status, and forum communication behavior into course participation status; and constructing a user behavior state model based on the classification results.

[0009] According to the above technical solution, the user behavior state model consists of course completion, course mastery, and community participation. Course completion refers to the number of video plays and the total playback time, course mastery refers to the simulated practice score, and community participation refers to the number of forum posts and comments.

[0010] For example, based on the total video playback time and the number of times the user plays the video, the following formula is defined to calculate course completion:

[0011]

[0012] Where num and NUM represent the number of times a user plays videos and the total number of course videos, respectively, and time and TIME represent the cumulative duration of video playback and the total duration of course videos, respectively.

[0013] For example, the following formula is defined to calculate course mastery based on the user's score for each simulated exercise:

[0014]

[0015] Where grade(i) represents the user's score in the i-th simulation practice session, and N represents the total number of simulation practice sessions the user has taken.

[0016] Specifically, based on the number of posts and comments made by users on the forum, the following formula for calculating community participation is defined:

[0017]

[0018] Among them, post(i) represents the influence factor of user forum posts, dis(j) represents the effective factor of user comments, and n and m respectively represent the number of user forum posts and the number of user comments;

[0019] Based on the above technical solution, for the n user behavior data of users who have completed all courses collected from the system background database, the analytic hierarchy process and its calculation formula are used to judge and measure the parameter weights of the model. The calculation formula of the analytic hierarchy process is as follows:

[0020]

[0021] Among them, ɑ ij represents the importance degree of the i-th element to the j-th element, and ɑ kj represents the importance degree of the k-th element to the j-th element;

[0022] Exemplarily, to obtain the parameter weights of the model, the following user behavior state model weight score formula is defined:

[0023] g(x) = ɑ * read + β * sco + γ * com

[0024] Among them, read is the defined course completion degree, sco is the defined course mastery degree, com is the defined community participation degree, and ɑ, β, and γ respectively represent the weights of the course completion degree, course mastery degree, and community participation degree.

[0025] Step S3: Use the user behavior state model to evaluate and judge the learning achievements of the user. The learning achievements of the user include the weight score of the user behavior state model and the course mastery degree;

[0026] According to the above technical solution, for the evaluation and judgment of the learning achievements of the user, based on the n user simulation practice data of users who have completed all courses collected from the system background database, the mastery degree and experience value algorithm formula is used to evaluate g(x) and sco. The mastery degree and experience value algorithm formula is as follows:

[0027]

[0028] Among them, degree is the minimum requirement for the user to master the course, norm is the normalization coefficient, g(x) i is the behavior state of the i-th user, sco i is the course mastery degree of the i-th user. When g(x) ≥ degree, the mastery degree of the user for the course knowledge points is marked as fully mastered. When g(x) < degree, the mastery degree of the user for the course knowledge points is marked as not fully mastered, and the mastery degree of the user for the learned course knowledge points is obtained

[0029] Step S4: Iteratively optimize the evaluation results of the user's learning outcomes using a dynamic adaptive function and a dynamic adaptive genetic algorithm to generate a personalized learning path.

[0030] According to the above technical solution, the dynamic adaptive function further includes: improving the fitness function in the original genetic algorithm, and using the improved dynamic adaptive function to calculate the real-time fitness score of the course;

[0031] The improved dynamic adaptive function formula is as follows:

[0032]

[0033] Where, θ(x) i ) is the real-time adaptability score of the i-th course. It is the real-time adaptability score of the i-th course before dynamic updates, g(x) i The degree represents the user's level of mastery over the i-th course. i This is the minimum requirement for a user to master the i-th course, δ(x) i ) represents the efficiency of a user learning the i-th course;

[0034] Furthermore, the dynamic fitness function formula is as follows:

[0035]

[0036] Where, p i Let θ(x) be the probability of recommending the i-th course. i f(x) is the real-time adaptability score of the i-th course. i ) is the recommendation value of the i-th course, and N is the total number of courses;

[0037] Finally, a dynamic adaptive genetic algorithm is used to iteratively optimize the learning progress of the user, and the optimized path is output as a personalized learning path recommendation result based on user behavior.

[0038] According to the above technical solution, the personalized learning path planning method based on user behavior is executed by a personalized learning path planning system based on user behavior, the system comprising the following modules:

[0039] The user learning behavior collection module is used to collect user learning behavior data from the system.

[0040] The user behavior state model building module is used to analyze and set the weight scores corresponding to different learning behaviors of users, and build a user behavior state model.

[0041] The user learning outcome analysis module is used to evaluate user learning outcomes and calculate the user's mastery of course knowledge points by using a user learning behavior and status model.

[0042] The personalized learning path planning module uses a dynamic adaptive genetic algorithm to iteratively optimize the user's learning path and dynamically update the recommended personalized learning path based on user behavior in real time.

[0043] According to the above technical solution, the present invention also discloses a computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the steps in the above-mentioned personalized learning path planning method based on user behavior.

[0044] According to the above technical solution, the present invention also discloses a computer device, including a memory, a processor, and computer instructions stored in the memory and running on the processor. When the processor runs the computer instructions, it implements the steps in the above-mentioned personalized learning path planning method based on user behavior.

[0045] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0046] (1) This invention collects the learning behaviors generated by users during the learning process, sets weight scores for different behaviors, and constructs a user behavior state model, which helps users understand the learning behaviors they generate and provides theoretical guidance for the planning and design of subsequent personalized learning paths.

[0047] (2) Integrate user learning behavior into the personalized learning path, adjust the learning path and content in a timely manner according to the user's learning progress and mastery of course knowledge points, improve the matching degree between users and courses, enhance user learning efficiency and experience, and ensure user learning quality.

[0048] (3) For the learning path optimization problem, the present invention adopts a dynamic adaptive genetic algorithm, which has the ability to search quickly and randomly without being independent of the problem domain. It can find the global optimal solution in all courses, and has a fast operation speed and strong robustness. It meets the requirements of real-time dynamic updates and can improve the accuracy of course recommendation. Attached Figure Description

[0049] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0050] In the attached diagram:

[0051] Figure 1 The present invention proposes a flowchart for a personalized learning path planning method based on user behavior;

[0052] Figure 2 This is a schematic diagram of the user behavior state model of the present invention;

[0053] Figure 3 This is a flowchart illustrating the personalized learning path planning process of this invention.

[0054] Figure 4 This invention proposes a system architecture diagram for a personalized learning path planning system based on user behavior. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] like Figure 1 As shown, this invention provides a personalized learning path planning method based on user behavior, including the following:

[0057] Step S1: Collect user learning behavior data.

[0058] When users are learning a chapter of a course, they will generate a series of learning behaviors, such as watching videos, doing mock exercises, and engaging in forum discussions. The system backend automatically records and updates the learning behaviors generated by users during the learning process.

[0059] Step S2: Construct a user behavior state model.

[0060] The learning behaviors collected in step S1 are categorized as follows: video viewing behavior is categorized as course completion status, simulation practice behavior is categorized as course mastery status, and forum communication behavior is categorized as course participation status.

[0061] Based on the classification results, a user behavior state model is constructed, such as... Figure 2 As shown, the user behavior state model consists of course completion, course mastery, and community participation. Course completion refers to the number of video views and the total viewing time, course mastery refers to the simulated practice score, and community participation refers to the number of forum posts and comments.

[0062] Based on the total duration of video playback and the number of times the user plays the video, a formula for calculating course completion is defined as follows:

[0063]

[0064] Where num and NUM represent the number of times a user plays videos and the total number of course videos, respectively, and time and TIME represent the cumulative duration of video playback and the total duration of course videos, respectively.

[0065] Based on the user's score in each simulated exercise, a formula is defined to calculate course mastery, as follows:

[0066]

[0067] Where grade(i) represents the user's score in the i-th simulation practice session, and N represents the total number of simulation practice sessions the user has taken.

[0068] Based on the number of posts and comments made by users on the forum, a formula for calculating community participation is defined as follows:

[0069]

[0070] Where post(i) represents the influence factor of user forum posts, dis(j) represents the effective factor of user comments, and n and m represent the number of user forum posts and the number of user comments, respectively;

[0071] Obtain behavioral data from n users who have completed all courses within the system platform. Use the Analytic Hierarchy Process (AHP) and its calculation formula to determine and measure the parameter weights of the model. The AHP calculation formula is as follows:

[0072]

[0073] Among them, α ij Let α represent the importance of the i-th element to the j-th element. kj This represents the importance of the k-th element to the j-th element.

[0074] Once the model's parameter weights are obtained, the formula for the weight score of the user behavior state model is defined as follows:

[0075] g(x) = α*read + β*sco + γ*com

[0076] Where read represents the defined course completion rate, sco represents the defined course mastery rate, com represents the defined community participation rate, and α, β, and γ represent the weights of course completion rate, course mastery rate, and community participation rate, respectively.

[0077] Step S3: Estimate and judge the user's learning outcomes.

[0078] Obtain simulated practice data from n users who have completed all courses from the system platform. Use the mastery and experience value algorithm to evaluate g(x) and sco in step S2. The formula for the mastery and experience value algorithm is as follows:

[0079]

[0080] Where degree is the minimum requirement for a user to complete the course, norm is the normalization coefficient, and g(x) is the minimum normalization coefficient. i For the behavioral state of the i-th user, sco i Let represent the course mastery level of the i-th user.

[0081] When g(x)≥degree, it is recorded that the user has fully mastered the knowledge points of this chapter; otherwise, it is recorded that the user has not fully mastered the knowledge points of this chapter, and finally the user's mastery of the knowledge points of the learned courses is obtained.

[0082] Step S4: Generate a personalized learning path.

[0083] Before the user takes the course for the first time, the system randomly generates a test paper from the question set and conducts a preliminary test to assess the user's understanding of the relevant knowledge points based on the accuracy of the questions, thus roughly forming the initial learning path.

[0084] From step S3, obtain the user's level of mastery of the knowledge points in the learned courses, such as... Figure 3 As shown, the learning path for the next stage of the course is determined based on the user's level of mastery of the course knowledge points. Specifically, if the user has not fully mastered the knowledge points of this chapter, the learned course will be reviewed again and related courses will be pushed to reinforce the knowledge. If the user has fully mastered the knowledge points of this chapter, the next stage of learning will begin.

[0085] The fitness function in the original genetic algorithm is improved, and the system obtains the real-time fitness score of the course through the improved dynamic adaptive function.

[0086] The improved dynamic adaptive function formula is as follows:

[0087]

[0088] Where, θ(x) i ) is the real-time adaptability score of the i-th course. It is the real-time adaptability score of the i-th course before dynamic updates, g(x) i The degree represents the user's level of mastery over the i-th course. i This is the minimum requirement for a user to master the i-th course, δ(x) i ) represents the efficiency of a user learning the i-th course.

[0089] By using a dynamic adaptive genetic algorithm and iteratively optimizing the learning progress of the user, the best-performing path is finally selected as the recommended learning path for the user.

[0090] Among them, the formula of the dynamic fitness function is as follows:

[0091]

[0092] Among them, p i is the recommendation probability of the i-th course, θ(x i ) is the real-time fitness score of the i-th course, f(x i ) is the recommendation value of the i-th course, and N is the total number of courses.

[0093] Example 2

[0094] Exemplarily, the background of the online training system collects the learning data of a certain user, obtains the learning behavior of the user for a certain course. The course completion degree read = 0.9, the course mastery degree sco = 0.8, and the community participation degree com = 0.75 of this user; based on the defined weight score formula of the user behavior state model g(x) = ɑ*read + β*sco + γ*com, where the course completion degree weight ɑ = 40, the course mastery degree weight β = 20, and the community participation degree weight γ = 20, calculate the mastery degree of the user for this course g(x) = 83; and obtain 100 user data of those who have completed learning this course from the system platform, and get the minimum requirement degree = 85 for the user to completely master this course. Because g(x) < degree, the system judges that the user has not completely mastered this course; the system adjusts the real-time fitness score θ(x) of this course according to the improved dynamic adaptive function, and allows the user to review the learned course again and push relevant courses for consolidation.

[0095] Example 3

[0096] As Figure 4 shown, this example proposes a personalized learning path planning system based on user behavior, including the following modules:

[0097] User learning behavior collection module: Collect user learning data from the training system background log, and obtain learning behaviors such as user video viewing, simulation practice, and forum communication;

[0098] User behavior state model construction module: Analyze and set the weight scores corresponding to different learning behaviors according to the user's learning behaviors, and construct a user behavior state model;

[0099] User learning result analysis module: Evaluate the user's learning achievements through the user's learning behaviors and state model, and estimate the user's mastery degree of course knowledge points;

[0100] Personalized learning path planning module: For course knowledge points where users are weak, an initial personalized learning path is generated. As the user's learning progresses, a dynamic adaptive genetic algorithm is used for iterative optimization, and the learning path is updated in real time. Knowledge points that have been fully mastered are deleted, and knowledge points that need to be fully mastered are added and recommended to the user.

[0101] Example 4

[0102] This embodiment provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the steps in the above-described personalized learning path planning method based on user behavior.

[0103] Example 5

[0104] This embodiment provides a computer device, including a memory, a processor, and computer instructions stored in the memory and running on the processor. When the processor runs the computer instructions, it implements the steps in the above-described personalized learning path planning method based on user behavior.

[0105] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0106] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0107] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1The function specified in one or more boxes.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0109] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0110] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A personalized learning path planning method based on user behavior, characterized in that: Includes the following steps: Step S1: Collect user learning behavior data based on the system's backend database. User behavior data includes video viewing data, simulation practice data, and forum communication data. Step S2: Classify the collected user learning behavior data and construct a user behavior state model based on the classification results. The user behavior state model consists of course completion rate, course mastery rate, and community participation rate. Define the following formula for the weight score of the user behavior state model: ； in, For the defined course completion level, For the defined course mastery level, For the community engagement defined, , , These represent the weights of course completion, course mastery, and community participation, respectively. Step S3: Use the user behavior state model to evaluate and judge the user's learning outcomes, which include the weight score of the user behavior state model and the course mastery. The learning outcomes of users are evaluated and judged based on simulated practice data from n users who have completed all courses, collected from the system's backend database. The evaluation is then performed using a mastery and experience value algorithm. and The evaluation, mastery level, and experience value algorithm formula are as follows: ； in, This sets the minimum requirements for users to master the course. The normalization coefficient is... Let i be the behavioral state of the i-th user. For the course mastery level of the i-th user, when At that time, the user's mastery of the course knowledge points is marked as complete. The time marker indicates that the user's mastery of the course knowledge points is not complete, thus obtaining the user's mastery level of the course knowledge points already learned; Step S4: Iteratively optimize the evaluation results of the user's learning outcomes using a dynamic adaptive function and a dynamic adaptive genetic algorithm to generate a personalized learning path; The dynamic adaptive function further includes: improving the fitness function in the original genetic algorithm, and using the improved dynamic adaptive function to calculate the real-time fitness score of the course; The improved dynamic adaptive function formula is as follows: ； in, It is the real-time adaptability score of the i-th course. It is the real-time adaptability score of the i-th course before dynamic updates. It represents the user's level of mastery over the i-th course. This is the minimum requirement for a user to master the i-th course. It represents the efficiency of a user learning the i-th course; Furthermore, the dynamic fitness function formula is as follows: ； in, It is the probability of recommending the i-th course. It is the real-time adaptability score of the i-th course. Here, N is the recommendation value for the i-th course, and N is the total number of courses. Finally, a dynamic adaptive genetic algorithm is used to iteratively optimize the learning progress of the user, and the optimized path is output as a personalized learning path recommendation result based on user behavior.

2. The personalized learning path planning method based on user behavior according to claim 1, characterized in that: The video viewing data further includes the number of times the video was played and the total playback time; the simulation practice data further includes the simulation practice score; and the forum communication data further includes the number of forum posts and comments.

3. The personalized learning path planning method based on user behavior according to claim 1, characterized in that: The process of classifying the collected user learning behavior data includes: classifying video viewing behavior into course completion status, simulated practice behavior into course mastery status, and forum communication behavior into course participation status. Based on the classification results, a user behavior state model is constructed.

4. The personalized learning path planning method based on user behavior according to claim 3, characterized in that: The user behavior state model consists of course completion, course mastery, and community participation. Course completion is the number of video plays and the total playback time. Course mastery is the simulated exercise score. Community participation is the number of forum posts and comments. Based on the total video playback time and number of plays, the following formula is defined to calculate course completion: ； Where num and NUM represent the number of times a user plays videos and the total number of course videos, respectively, and time and TIME represent the cumulative duration of video playback and the total duration of course videos, respectively. Based on the user's score for each simulated exercise, the following formula is defined to calculate course mastery: ； in, Let N represent the user's score in the i-th simulation practice session, and N represent the total number of simulation practice sessions the user has taken. Furthermore, based on the number of posts and comments made by users on the forum, the following formula is defined for calculating community participation: ； in, This indicates the influence factor of user forum posts. The effective factor represents user comments, where n and m represent the number of forum posts and comments made by the user, respectively. Based on the behavior data of n users who have completed all courses, collected from the system's backend database, the analytic hierarchy process (AHP) and its calculation formula are used to determine and measure the parameter weights of the model. The AHP calculation formula is as follows: ； in, To represent the importance of the i-th element to the j-th element, This represents the importance of the k-th element to the j-th element; Obtain the parameter weights of the model.

5. The personalized learning path planning method based on user behavior according to claim 1, executed by a personalized learning path planning system based on user behavior, characterized in that: The system includes the following modules: The user learning behavior collection module is used to collect user learning behavior data from the system. The user behavior state model building module is used to analyze and set the weight scores corresponding to different learning behaviors of users, and build a user behavior state model. The user learning outcome analysis module is used to evaluate user learning outcomes and calculate the user's mastery of course knowledge points by using a user learning behavior and status model. The personalized learning path planning module uses a dynamic adaptive genetic algorithm to iteratively optimize the user's learning path and dynamically update the recommended personalized learning path based on user behavior in real time.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the steps of the personalized learning path planning method based on user behavior according to any one of claims 1-4.

7. A computer device, characterized in that: The computer device includes a memory, a processor, and computer instructions stored in the memory and running on the processor. When the processor executes the computer instructions, it implements the steps of the personalized learning path planning method based on user behavior according to any one of claims 1-4.

Images