Dynamic learning resource recommendation system driven by intelligent behavior analysis teaching-assistant machine

Abstract

The invention relates to the technical field of dynamic learning resource recommendation. The invention relates to a dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching aid. The system comprises a resource library establishment unit, an initial resource updating unit, an examination prediction unit, a resource selection unit and a resource stage management unit. The resource library establishing unit is used for establishing a resource library in the teaching assistant machine, and collecting grade information of each student, teaching progress of a teacher and examination data of the student at the same time; by calculating the knowledge point matching degree of the wrong questions and the learning resources, it is ensured that knowledge points corresponding to the wrong questions can be matched with related content in a resource library, accurate mapping is formed, low learning efficiency caused by generalization recommendation is avoided, by integrating multi-dimensional data portrait support, the recommendation resources are made to accurately match individual demands of students, and the learning efficiency is improved. Resource difficulty is prevented from being disjointed with student level.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of dynamic learning resource recommendation, in particular to a dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant machine. BACKGROUND

[0002] In the field of intelligent education, the existing technology usually constructs a recommendation model based on single-dimensional data such as grade or test scores, and realizes resource pushing through rule matching or simple collaborative filtering algorithm. The core purpose is to solve the problem of blindness in resource distribution in traditional education and preliminarily realize the weak association matching of learning resources and students' foundation.

[0003] Its working logic depends on fixed templates or historical data statistics, and only classifies and recommends resources according to the students' grades and subjects, or generates unified review content based on the correct answer rate of the group, which lacks real-time tracking of students' individual learning behavior, resulting in recommended resources that cannot dynamically adapt to the changes in learning ability during the learning process. In order to reduce this situation, a dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant machine is proposed. SUMMARY

[0004] The purpose of the present application is to provide a dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant machine to solve the problems raised in the background art.

[0005] To achieve the above purpose, a dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant machine is provided, which includes a resource library establishment unit, an initial resource updating unit, a test prediction unit, a resource selection unit and a resource stage management unit.

[0006] The resource library establishment unit is used to establish a resource library in the teaching assistant machine, and collect the grade information of each student, the teaching progress of the teacher and the test data of the student at the same time.

[0007] The initial resource updating unit is used to complete the initial recommendation of learning resources in the resource library for each student's autonomous learning according to the interannual information and the teaching progress, and record the learning behavior of the student's autonomous learning, and update the learning resources across the span using the learning behavior record.

[0008] The test prediction unit is used to correctly predict the questions contained in the test data, and then extract the wrong question data according to the correct prediction combined with the test data.

[0009] The resource selection unit is used to analyze the knowledge points of the wrong question data and the resource library at the same time, match the learning resources of the same knowledge points, and then select the representative learning resources for each knowledge point and the proportion of the knowledge points according to the wrong question data.

[0010] The resource stage management unit is used for acquiring updated difference data of teaching progress, performing learning resource proportion allocation according to the updated difference data in combination with the wrong question data, and then performing stage updating on the learning resource according to the allocation result.

[0011] As a further improvement of the technical solution, the resource library establishing unit collects learning resources online through the Internet and saves the learning resources in the teaching assistant machine.

[0012] The grade information is acquired by collecting the personal information of the students.

[0013] The teaching progress is acquired by collecting the teaching diaries of the teachers and performing analysis according to the teaching diaries.

[0014] The examination data of each student is collected by collecting the examination paper data.

[0015] As a further improvement of the technical solution, the initial resource updating unit comprises an initial recommendation module and a span updating module.

[0016] The initial recommendation module is used for allocating a corresponding teacher to each student, performing adaptive learning range analysis on the teaching progress of the corresponding teacher in combination with the grade information of the student, performing autonomous learning initial recommendation analysis on the adaptive learning range in combination with the resource library, and initially recommending autonomous learning corresponding learning resources to the student according to the analysis result.

[0017] The span updating module is used for monitoring learning behavior records of the student in autonomous learning driven by the teaching assistant machine, performing learning quality analysis on each newly appeared learning behavior record, and performing span updating on the learning resources in combination with the learning resources initially recommended.

[0018] As a further improvement of the technical solution, the span updating module adds an interactive process to the learning resources in the autonomous learning process of the student, collects learning feeling of the student according to the interactive process, and then performs learning quality analysis on the learning feeling and learning time length.

[0019] As a further improvement of the technical solution, the examination prediction unit comprises a question analysis module and a wrong question extraction module.

[0020] The question analysis module is used for extracting answering data and question data from the examination data, performing correct prediction on the question data in combination with the learning behavior records and the teaching progress, and acquiring predicted questions of the student.

[0021] The wrong question extraction module is used for performing wrong question analysis on the question data and the answering data, acquiring questions answered incorrectly by the student, and then performing same question comparison between the questions answered incorrectly and the predicted questions, and collecting the same questions as the wrong question data.

[0022] As a further improvement of the technical solution, the resource selection unit comprises a knowledge point analysis module and a representative selection module.

[0023] The knowledge point analysis module is configured to analyze the relevant knowledge points of each question in the wrong question data, and analyze the relevant knowledge points of each learning resource in the resource library, so as to obtain the relevant knowledge points corresponding to each question and the relevant knowledge points corresponding to each learning resource.

[0024] The representative selection module is configured to match the relevant knowledge points of the wrong question data with the relevant knowledge points of the learning resources, match the learning resources with the same knowledge points as the wrong question data, and then select the matched learning resources as the representative learning resources for each knowledge point and knowledge point proportion based on the wrong question data, obtain the proportion values of the wrong question data for different learning resources, and then select one representative learning resource for each knowledge point.

[0025] As a further improvement of the technical solution, when calculating the proportion value, the more the number of knowledge points related to a teaching resource is the same as the number of knowledge points related to the wrong question data, the higher the proportion value is.

[0026] Meanwhile, when selecting the representative learning resources, the selected representative learning resources cover all the relevant knowledge points of the wrong question data.

[0027] As a further improvement of the technical solution, the resource stage management unit comprises a proportion analysis module and a stage updating module.

[0028] The proportion analysis module is configured to analyze the difference between the teaching progress before the update and the teaching progress after the update when the teaching progress is updated, so as to obtain the update difference data corresponding to the teacher's teaching.

[0029] The proportion analysis module is configured to analyze the difference between the teaching progress before the update and the teaching progress after the update when the teaching progress is updated, so as to obtain the update difference data corresponding to the teacher's teaching.

[0030] The stage updating module is configured to analyze the recommended learning resources based on the weights of the update difference data and the wrong question data, and then send the recommended learning resources to the teaching assistant machine to complete the learning resource stage update for the students.

[0031] As a further improvement of the technical solution, when the proportion weight is allocated, the higher the proportion value in the wrong question data prediction question is, the higher the weight is, and vice versa.

[0032] The smaller the difference in the updated difference data, the lower the weight. Conversely, the larger the difference in the updated difference data, the higher the weight.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. In this dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant, the matching degree between wrong questions and knowledge points in learning resources is calculated to ensure that the knowledge points corresponding to each wrong question can be matched to the relevant content in the resource library, forming an accurate mapping, avoiding the low learning efficiency caused by generalized recommendations. By integrating multi-dimensional data portrait support, the recommended resources are accurately matched to the individual needs of students, avoiding the disconnection between resource difficulty and student level.

[0035] 2. In this dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant, by predicting the questions that students can answer correctly and combining the actual wrong questions to screen out the questions that "should have been mastered but answered incorrectly", a personalized wrong question set is generated. This mechanism prevents students from blindly practicing questions and focuses on their real weaknesses. Questions that are predicted to be correct but actually answered incorrectly are used as key review content to improve review efficiency. By giving priority to resources that cover all knowledge points of wrong questions and have a high proportion, it ensures that each knowledge point corresponds to a unique high-quality resource to avoid repeated recommendations. If a knowledge point appears in multiple wrong questions, it is automatically matched with the content in the resource library that covers the knowledge point and has the highest matching degree to form a structured learning path. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is the overall structural principle diagram of the present invention.

[0037] The meaning of each number in the figure is:

[0038] 10. Resource library establishment unit; 20. Initial resource update unit; 30. Test prediction unit; 40. Resource selection unit; 50. Resource stage management unit. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0040] See also Figure 1 As shown, the present embodiment aims to provide a dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant, comprising a resource library establishment unit 10, an initial resource update unit 20, an exam prediction unit 30, a resource selection unit 40, and a resource stage management unit 50;

[0041] The resource library establishment unit 10 is used to establish a resource library on the teaching assistant machine, and collect each student's grade information, the teacher's teaching progress and the student's test data;

[0042] The resource library establishment unit 10 collects learning resources online through the Internet and stores the learning resources in the teaching assistant machine;

[0043] Use web crawlers to grab resources, standardize and label the resource formats, and then store them in the local database of the teaching assistant machine;

[0044] Obtain grade information by collecting students' personal information;

[0045] Parse registration information and student data.

[0046] By collecting teachers' teaching diaries and analyzing them, we can obtain the teaching progress;

[0047] This is accomplished through semantic analysis of teaching diaries and generation of teaching progress vectors.

[0048] By collecting test paper data, the test data of each student is collected.

[0049] The examination can be a monthly examination, a midterm examination, or a weekly examination. After each examination, images of the examination papers are collected and OCR recognition is performed.

[0050] The initial resource updating unit 20 is used to complete the initial recommendation of learning resources in the resource library for each student's independent learning based on inter-year information and teaching progress, and record the student's independent learning behavior and use the learning behavior record to update the learning resources.

[0051] The initial resource updating unit 20 includes an initial recommendation module and a span updating module;

[0052] The initial recommendation module is used to assign a teacher to each student. The module analyzes the teacher's teaching progress and the student's grade level to determine their learning scope. The module then analyzes the learning scope and resource library to perform an initial recommendation analysis for self-study. Based on the analysis results, the module recommends learning resources for self-study to the student. The specific steps are as follows:

[0053] Teacher and student assignment: Each student is assigned to a specific teacher based on grade information, and each student has a corresponding teacher;

[0054] Student learning scope analysis: Determine each student's learning scope, consider their grade level, knowledge level, and individual needs, and select appropriate learning content;

[0055] Filtering suitable resources from the repository: From the learning resource library, filter out resources that meet the needs of students and teaching progress, each resource will contain relevant knowledge points and difficulty levels;

[0056] Priority ranking and resource recommendation: According to the student's historical learning data, teacher feedback and resource relevance, sort and filter out the best resources for recommendation;

[0057] The span update module is used to monitor the student's learning behavior record driven by the teaching machine for autonomous learning, and then analyze the learning quality of each new learning behavior record. The learning quality is combined with the initial recommended learning resources for span update.

[0058] The span update module adds an interactive process to the learning resources during the student's autonomous learning process, collects learning feelings from students according to the interactive process, and then summarizes learning feelings and learning duration for learning quality analysis. The specific steps are as follows:

[0059] Monitor student learning behavior: The teaching machine records various learning behavior data in real time during the student's autonomous learning process, such as: learning duration, progress, review frequency, correct answer rate, task completion, learning content selection, etc.

[0060] Analysis and quality evaluation of learning behavior records: Learning quality analysis is performed on each new learning behavior record, taking into account the following aspects:

[0061] Knowledge mastery level: assess the knowledge mastery level through the student's task completion, test scores, content review, etc.

[0062] Learning effect evaluation: evaluate the relationship between learning duration, concentration and task completion to determine the student's learning effect

[0063] Learning progress comparison: compare the student's current learning progress with the initial learning plan to evaluate the learning progress;

[0064] Evaluation and span update of initial recommended learning resources: According to the student's learning foundation, knowledge mastery level and interest preference, the system recommends a set of initial learning resources for the student. Then, with the record and analysis of the student's learning behavior, the learning effect of these initial recommended resources will be evaluated, combined with the student's learning quality score, to determine whether the recommended resources need to be updated or adjusted to better match the student's learning needs and progress. The formula is as follows:

[0065] Q = w1 × Ta co + w2 × Le tm + w3 × Re fe + w4 × Kn dg ;

[0066] Wherein, Q is the learning quality score, Ta co is the student task completion degree, Le tm is the student learning time, Re fe is the review frequency, Kn dg is the student retention of learning content, w1, w2, w3, w4 are weight coefficients.

[0067] The test prediction unit 30 is used for correct prediction of the questions contained in the test data, and then wrong question data extraction is carried out according to the correct prediction combined with the test data;

[0068] The test prediction unit 30 includes a question analysis module and a wrong question extraction module;

[0069] The question analysis module is used for extracting the answer data (including the answers selected by the student in the test and whether the answer is correct) and the question data (including the content, type and belonging knowledge point of each question) of the test data, combining the question data with the learning behavior record and the teaching progress to make correct prediction, and obtaining the predicted questions that the student can answer correctly;

[0070] Based on the learning behavior record, knowledge point mastery and teaching progress of the student, a prediction model is constructed to predict the questions that the student can answer correctly;

[0071] The model input is the learning behavior record of the student, the question data and the teaching progress;

[0072] The model output is the question set that the student can answer correctly, and the formula is as follows:

[0073]

[0074] Wherein, P(T|K) is the probability that the student answers the question T correctly after mastering the knowledge point K, P(K|T) is the probability that the student masters the knowledge point K given the question T, P(T) is the prior probability of the question T (i.e. the probability that all students answer the question correctly), and P(K) is the prior probability that the student masters the knowledge point K.

[0075]

[0076] Wherein, n is the number of knowledge points, a i is the weight of each knowledge point K i to the question T, P(T|K i ) is the probability that the student answers the question T correctly under the condition that the knowledge point K i .

[0077] The wrong question extraction module is used for wrong question analysis of the question data and the answer data, obtaining the questions answered incorrectly by the student, and then comparing the questions answered incorrectly with the predicted questions, and summarizing the questions with the same predicted questions as the wrong question data.

[0078] By comparing the questions predicted to be answered correctly by the student with the questions actually answered incorrectly by the student, it is found out which predicted questions are actually answered incorrectly by the student:

[0079] The questions predicted to be answered correctly by the student and the questions answered incorrectly by the student are obtained, and the two are compared to find out which questions are in both the predicted question list and the wrong question list, and these questions are summarized as wrong question data for subsequent personalized learning plan.

[0080] Finally, after comparing the predicted questions with the wrong questions, a wrong question set is generated, and the wrong questions are summarized, and the wrong question data structure is the student identity and the wrong question set.

[0081] The resource selection unit 40 is used for knowledge point analysis of the wrong question data and the resource library at the same time, and matching of learning resources of the same knowledge point, and then selecting a representative learning resource for each knowledge point and knowledge point proportion according to the wrong question data;

[0082] The resource selection unit 40 includes a knowledge point analysis module and a representative selection module;

[0083] The knowledge point analysis module is used for related knowledge point analysis of each question of the wrong question data, and related knowledge point analysis of each learning resource in the resource library, so as to obtain the corresponding related knowledge point of each question and the corresponding related knowledge point of each learning resource;

[0084] The representative selection module is used for knowledge point matching according to the corresponding related knowledge point of the wrong question data combined with the corresponding related knowledge point of the learning resource, matching to obtain the learning resource of the same knowledge point of the wrong question data, and then selecting a representative learning resource for each knowledge point and knowledge point proportion combined with the wrong question data, obtaining the proportion value of the wrong question data for different learning resources, and then selecting a representative learning resource for each knowledge point.

[0085] The knowledge point set of each wrong question and the knowledge point set of each learning resource can be matched, and the matching rule is that if the wrong question and the learning resource have one or more same knowledge points, they are considered to match;

[0086]

[0087] Wherein, M(CT i , R j ) is the matching degree of the wrong question and the learning resource, CK iis the set of knowledge points corresponding to the wrong question data, RL j For each learning resource, its corresponding knowledge point set, |CK i ∩RL j | is the size of the intersection of the knowledge points of the wrong questions and the learning resources, that is, the number of knowledge points they contain together, |CK i ∪RL j | is the size of the union of the knowledge points of wrong questions and learning resources, that is, the total number of knowledge points they have.

[0088] When the representative selection module calculates the percentage value, the more the number of knowledge points related to a teaching resource is the same as the number of knowledge points related to the wrong question data, the higher the percentage value;

[0089] At the same time, when selecting representative learning resources, the selected representative learning resources cover all relevant knowledge points of the wrong question data.

[0090] In order to ensure that the representative learning resources can cover all relevant knowledge points, it is necessary to select a unique representative learning resource for each knowledge point. These learning resources should cover all relevant knowledge points in the wrong question data, and select the most appropriate learning resource for each knowledge point.

[0091] The resource stage management unit 50 is used to obtain updated difference data of the teaching progress, allocate learning resources in proportion according to the updated difference data combined with wrong question data, and then update the learning resources in stages according to the allocation results.

[0092] The resource phase management unit 50 includes a ratio analysis module and a phase update module;

[0093] The proportion analysis module is used to analyze the teaching progress before and after the update when the teaching progress is updated, so as to obtain the update difference data corresponding to the teacher's teaching.

[0094] When the teaching schedule is updated, analyze and compare the content of knowledge points (compare the teaching outlines before and after the update to identify new, deleted, or modified knowledge points) and the differences in time allocation (analyze the changes in the schedule before and after the update, for example, whether some knowledge points are allocated more teaching time, or whether some parts are compressed);

[0095] To compare differences, you can add new knowledge points, delete knowledge points, and modify content.

[0096] Combine the updated difference data with the wrong question data to perform a weighted allocation of learning resources, and obtain the weights of the updated difference data and the wrong question data when updating the learning resources;

[0097] The proportion analysis module, when performing proportion weight distribution, the higher the proportion of wrong question data in predicting the question, the higher the weight, and vice versa, the lower the proportion of wrong question data in predicting the question, the lower the weight.

[0098] When the difference in the update difference data is smaller, the weight is lower, and vice versa, when the difference in the update difference data is larger, the weight is higher.

[0099] W final (K i ) = beta * W err (K i ) + epsilon * W diff (K i );

[0100] Wherein, W final (K i ) is the maximum weight, combining the wrong question data weight and the update difference data weight, W err (K i ) is the weight of the wrong question data, W diff (K i ) is the weight of the update difference data, beta and epsilon are weight coefficients, indicating the relative importance of the wrong question data and the update difference data in the calculation of the maximum weight.

[0101] The stage update module is used for learning resource recommendation analysis according to the weights of the update difference data and the wrong question data, and then sends the recommended learning resources to the teaching assistant machine to complete the learning resource stage update of the student.

[0102] Through comprehensive analysis of the personal information, learning behavior, teaching progress of teachers and examination data of students, accurate learning resource recommendation can be provided for each student, helping students to continuously adjust and optimize learning content in the process of autonomous learning, and dynamic update can be performed according to real-time data, so that the recommended learning resources can always adapt to the learning needs of students.

[0103] The above shows and describes the basic principles, main features and advantages of the present application. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only preferred examples of the present application, and are not intended to limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant, characterized by: It includes a resource library establishment unit (10), an initial resource update unit (20), an examination prediction unit (30), a resource selection unit (40) and a resource stage management unit (50); The resource library establishment unit (10) is used to establish a resource library on the teaching assistant machine, and collect the grade information of each student, the teacher's teaching progress and the student's test data; The initial resource updating unit (20) is used to complete the initial recommendation of learning resources in the resource library for each student's autonomous learning based on grade information and teaching progress, and at the same time record the student's autonomous learning behavior and use the learning behavior record to update the learning resources in a span; The test prediction unit (30) is used to correctly predict the questions included in the test data, and then extract wrong question data based on the correct prediction and the test data; The resource selection unit (40) is used to simultaneously analyze the knowledge points of the wrong question data and the resource library, match learning resources with the same knowledge points, and then select representative learning resources for each knowledge point and the knowledge point ratio according to the wrong question data; The resource stage management unit (50) is used to obtain updated difference data of the teaching progress, allocate learning resources in proportion according to the updated difference data combined with wrong question data, and then update the learning resources in stages according to the allocation results.

2. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 1, characterized in that: The resource library establishment unit (10) collects learning resources online through the Internet and stores the learning resources in the teaching assistant machine; Obtain grade information by collecting students' personal information; By collecting teachers' teaching diaries and analyzing them, we can obtain the teaching progress; By collecting test paper data, the test data of each student is collected.

3. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 1, characterized in that: The initial resource updating unit (20) includes an initial recommendation module and a span updating module; The initial recommendation module is used to assign a corresponding teacher to each student, analyze the teacher's teaching progress in combination with the student's grade information to determine the learning range, and then analyze the learning range in combination with the resource library to perform an initial recommendation analysis for self-study. Based on the analysis results, the module initially recommends learning resources corresponding to self-study to the student. The span update module is used to monitor the learning behavior records of students in autonomous learning driven by the teaching assistant, and then analyze the learning quality of each new learning behavior record, and update the span based on the learning quality and the initially recommended learning resources.

4. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 3, characterized in that: The span updating module adds an interactive process to the learning resources during the students' autonomous learning process, collects students' learning feelings according to the interactive process, and then summarizes the learning feelings and learning time to analyze the learning quality.

5. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 1, characterized in that: The test prediction unit (30) includes a question analysis module and a wrong question extraction module; The question analysis module is used to extract answer data and question data from the test data, and accurately predict the question data by combining it with the learning behavior record and teaching progress to obtain the predicted question that the student will answer correctly; The wrong question extraction module is used to analyze the question data and the answer data to obtain the questions that the students answered incorrectly, and then compare the wrongly answered questions with the predicted questions for the same questions, and summarize the questions that are the same as the predicted questions as the wrong question data.

6. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 1, characterized in that: The resource selection unit (40) includes a knowledge point analysis module and a representative selection module; The knowledge point analysis module is used to analyze the relevant knowledge points of each question in the wrong question data, and also to analyze the relevant knowledge points of each learning resource in the resource library, so as to obtain the relevant knowledge points corresponding to each question and the relevant knowledge points corresponding to each learning resource; The representative selection module is used to match knowledge points according to the relevant knowledge points corresponding to the wrong question data and the relevant knowledge points corresponding to the learning resources, match and obtain learning resources with the same knowledge points as the wrong question data, and then combine the matched learning resources with the wrong question data to select representative learning resources for each knowledge point and the proportion of knowledge points, obtain the proportion value of different learning resources to the wrong question data, and then select a representative learning resource for each knowledge point.

7. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 1, characterized in that: When the representative selection module calculates the proportion value, the more the number of knowledge points related to a teaching resource is the same as the number of knowledge points related to the wrong question data, the higher the proportion value; At the same time, when selecting representative learning resources, the selected representative learning resources cover all relevant knowledge points of the wrong question data.

8. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 1, characterized in that: The resource phase management unit (50) includes a ratio analysis module and a phase update module; The ratio analysis module is used to analyze the teaching progress before and after the update when the teaching progress is updated, so as to obtain the update difference data corresponding to the teacher's teaching. Combine the updated difference data with the wrong question data to perform a weighted allocation of learning resources, and obtain the weights of the updated difference data and the wrong question data when updating the learning resources; The stage update module is used to perform learning resource recommendation analysis based on the weights corresponding to the updated difference data and the wrong question data, and then send the recommended learning resources to the teaching assistant machine to complete the stage update of the student's learning resources.

9. The dynamic learning resource recommendation system driven by an intelligent behavior analysis teaching assistant according to claim 8, characterized in that: When the proportion analysis module allocates proportion weights, the higher the proportion of wrong question data in the predicted questions, the higher the weight; conversely, the lower the proportion of wrong question data in the predicted questions, the lower the weight; The smaller the difference in the updated difference data, the lower the weight. Conversely, the larger the difference in the updated difference data, the higher the weight.

Images