Education resource retrieval method and device based on cognitive diagnosis and learning path

Abstract

The application discloses an education resource retrieval method and device based on cognitive diagnosis and learning path, and the method comprises the following steps: determining a test question-knowledge point matrix and a user-test question matrix according to a knowledge point set, and inputting the matrix into a cognitive diagnosis model, so that the model judges the comprehensive mastery degree of each user on each knowledge point, and obtains weak knowledge points of the user; the test question-knowledge point matrix comprises a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix; and the user-test question matrix comprises a user-video matrix, a user-practice question matrix and a user-other resource matrix. The AprioriAll algorithm is used to scan the knowledge point set, the weight value of the weak knowledge points of the user is increased in the learning path planning process, and a recommended learning path is generated; a plurality of member search engines are introduced to perform resource retrieval on the recommended learning path and integration on retrieval results, so that a recommended result is obtained, and the accuracy of the education resource retrieval is improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of cognitive diagnosis method, and particularly relates to an education resource retrieval method and device based on cognitive diagnosis and learning path. BACKGROUND

[0002] With the rapid development of Internet technology and the wide application of computer technology, human beings have entered an era of information explosion. Learners no longer limit the way of obtaining knowledge to books and classrooms, but shift more energy to online learning. However, when learners face a large amount of education resources, they often fall into the problems of "information overload" and "resource wandering", which reduces their learning efficiency and interest. Therefore, how to recommend personalized retrieval content for learners to meet their learning needs is a research hotspot in the field of education.

[0003] Under the network learning scenario, the key to overcoming the two problems of "learning wandering" and "cognitive overload" is how to help learners filter out resources suitable for themselves from the growing learning resources, and organize learning resources into a coherent learning path according to the relationship between the learning interests, cognitive level of learners and learning resources, so as to help learners improve learning efficiency. Therefore, under the trend of "Internet + education", constructing a personalized learning path recommendation system to adapt to the cognitive differences of different learners has become a hot issue for many scholars to study. However, the current learning resource recommendation has the following two outstanding problems:

[0004] (1) The mainstream way of learning path recommendation is to recommend completely according to the interests of learners, ignoring the personalized weak knowledge points of learners. However, unlike the recommendation idea according to user preferences in other fields, the real meaning of the recommendation system in the field of intelligent education should be to help learners quickly identify and make up for personalized weak knowledge points. The mainstream method cannot help learners check and make up for the weak points and effectively improve the learning efficiency.

[0005] (2) The mainstream way of weak knowledge point diagnosis is to make a judgment on the mastery of knowledge points of learners according to the exercise data. However, with the spread of streaming media videos, the mainstream method only considers judging from the exercise aspect, and cannot obtain an accurate knowledge point mastery.

[0006] However, in the face of today's education field, the existing cognitive diagnosis method cannot accurately obtain the weak knowledge points of the learners, and using the traditional DINA model (Deterministic Inputs, Noisy "And" gate model, a typical discrete cognitive diagnosis model) for cognitive diagnosis cannot discretize the mastery degree of the learners on the knowledge points into a continuous interval of [0, 1] due to the use of a two-level scoring mechanism. Moreover, the traditional learning path recommendation algorithm often has a cold start problem and does not take into account the learning situation of the learners.

[0007] Therefore, how to accurately obtain the weak knowledge points of the learners to generate a suitable recommended learning path for the learners and obtain a more accurate education resource retrieval result has become an important problem. SUMMARY

[0008] In order to solve the above problems existing in the prior art, the present application provides an education resource retrieval method and device based on cognitive diagnosis and learning path.

[0009] The technical problem to be solved by the present application is solved by the following technical scheme:

[0010] In a first aspect, the present application provides an education resource retrieval method based on cognitive diagnosis and learning path, which comprises:

[0011] determining a question-knowledge point matrix and a user-question matrix according to a knowledge point set; the question-knowledge point matrix comprises a video-knowledge point matrix, a practice question-knowledge point matrix and an other resource-knowledge point matrix; the user-question matrix comprises a user-video matrix, a user-practice question matrix and a user-other resource matrix; the question-knowledge point matrix is used to represent the association degree of questions and knowledge points; the user-question matrix is used to represent the answering situation of the user to the questions;

[0012] inputting the question-knowledge point matrix and the user-question matrix into a cognitive diagnosis model, so that the cognitive diagnosis model obtains the weak knowledge points of the users by judging the comprehensive mastery degree of each user on each knowledge point; the weak knowledge points of the users are the knowledge points with a comprehensive mastery degree less than a preset threshold value;

[0013] scanning the knowledge point set using an AprioriAll algorithm, combining the weak knowledge points of the users, and generating a recommended learning path by increasing the weight value of the weak knowledge points of the users in the learning path planning process;

[0014] introducing a plurality of member search engines to perform resource retrieval on the recommended learning path, integrating the retrieval results under each member search engine, and obtaining a recommended result.

[0015] Optionally, the cognitive diagnosis model comprises a VFuzzy-CDF model, a CFuzzy-CDF model and a Fuzzy-CDF model, and the test question-knowledge point matrix and the user-test question matrix are input into the cognitive diagnosis model, so that the cognitive diagnosis model obtains the user weak knowledge points by judging the comprehensive mastery degree of each user on each knowledge point, comprising:

[0016] The video-knowledge point matrix and the user-video matrix are input into the VFuzzy-CDF model, so that the VFuzzy-CDF model obtains the user first weak knowledge points by judging the comprehensive mastery degree of each user on each knowledge point;

[0017] The other resource-knowledge point matrix and the user-other resource matrix are input into the CFuzzy-CDF model, so that the CFuzzy-CDF model obtains the user second weak knowledge points by judging the comprehensive mastery degree of each user on each knowledge point;

[0018] The exercise-knowledge point matrix and the user-exercise matrix are input into the Fuzzy-CDF model, so that the Fuzzy-CDF model obtains the user third weak knowledge points by judging the comprehensive mastery degree of each user on each knowledge point;

[0019] The user first weak knowledge points, the user second weak knowledge points and the user third weak knowledge points are fused to obtain the user weak knowledge points.

[0020] Optionally, the test question-knowledge point matrix and the user-test question matrix are input into the cognitive diagnosis model, so that the cognitive diagnosis model obtains the user weak knowledge points by judging the comprehensive mastery degree of each user on each knowledge point in the following manner:

[0021] According to the test question-knowledge point matrix, the fuzzy attribute weight of each test question on each knowledge point is obtained; the fuzzy attribute weight is used to represent the correlation degree of each test question and each knowledge point;

[0022] According to the fuzzy attribute weight of each test question on each knowledge point and the mastery degree of each user on each knowledge point, the mastery degree of each user on the required knowledge points in each test question is obtained;

[0023] Based on the mastery degree of each user on the required knowledge points in each test question, the actual score rate of each user is obtained by introducing the error rate and the guess rate of the user;

[0024] The actual score rate of each user and the correct answer probability of the user to each test question in the test question-knowledge point matrix are combined to determine the comprehensive mastery degree of each knowledge point by the user, and the knowledge point with a comprehensive mastery degree less than a preset threshold is taken as a weak knowledge point of the user.

[0025] Optionally, a plurality of member search engines are introduced to search resources for the recommended learning path, and the search results under each member search engine are integrated to obtain a recommended result, including:

[0026] A plurality of member search engines are introduced to search resources for the recommended learning path, and the search results under each member search engine are integrated to obtain a primary search result.

[0027] After invalid search results are removed from the primary search result, the remaining search results are sorted to obtain a recommended result.

[0028] Optionally, after invalid search results are removed from the primary search result, the remaining search results are sorted to obtain a recommended result, including:

[0029] After invalid search results are removed from the primary search result, the remaining search results are sorted in combination with the accuracy, recall rate and nDCG of each member search engine to obtain a recommended result.

[0030] In a second aspect, the present application provides an educational resource retrieval device based on cognitive diagnosis and learning path, which comprises:

[0031] A matrix determination module is configured to determine a test question-knowledge point matrix and a user-test question matrix according to a knowledge point set; the test question-knowledge point matrix includes a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix; the user-test question matrix includes a user-video matrix, a user-practice question matrix and a user-other resource matrix; the test question-knowledge point matrix is used to represent the correlation degree between test questions and knowledge points; and the user-test question matrix is used to represent the answering situation of the user to the test questions.

[0032] A weak knowledge point determination module is configured to input the test question-knowledge point matrix and the user-test question matrix into a cognitive diagnosis model, so that the cognitive diagnosis model determines the comprehensive mastery degree of each user to each knowledge point to obtain a weak knowledge point of the user; the weak knowledge point of the user is a knowledge point with a comprehensive mastery degree less than a preset threshold.

[0033] A recommended learning path generation module is configured to use an AprioriAll algorithm to scan the knowledge point set, combine the weak knowledge point of the user, and generate a recommended learning path by increasing the weight value of the weak knowledge point of the user in the learning path planning process.

[0034] The retrieval result integration module is configured to introduce a plurality of member search engines to search for resources of the recommended learning path, and integrate the retrieval results of each member search engine to obtain a recommended result.

[0035] Optionally, the cognitive diagnosis model includes a VFuzzy-CDF model, a CFuzzy-CDF model and a Fuzzy-CDF model, the user weak knowledge point determination module is specifically configured to input the video-knowledge point matrix and the user-video matrix into the VFuzzy-CDF model, so that the VFuzzy-CDF model determines the comprehensive mastery degree of each user for each knowledge point to obtain a first weak knowledge point of the user; input the other resource-knowledge point matrix and the user-other resource matrix into the CFuzzy-CDF model, so that the CFuzzy-CDF model determines the comprehensive mastery degree of each user for each knowledge point to obtain a second weak knowledge point of the user; input the exercise-knowledge point matrix and the user-exercise matrix into the Fuzzy-CDF model, so that the Fuzzy-CDF model determines the comprehensive mastery degree of each user for each knowledge point to obtain a third weak knowledge point of the user; and fuse the first weak knowledge point of the user, the second weak knowledge point of the user and the third weak knowledge point of the user to obtain the user weak knowledge point.

[0036] Optionally, the user weak knowledge point determination module inputs the question-knowledge point matrix and the user-question matrix into the cognitive diagnosis model, so that the cognitive diagnosis model determines the comprehensive mastery degree of each user for each knowledge point in the following manner to obtain the user weak knowledge point:

[0037] obtain the fuzzy attribute weight of each question for each knowledge point according to the question-knowledge point matrix; the fuzzy attribute weight is used to represent the correlation degree between each question and each knowledge point;

[0038] obtain the mastery degree of each user for the required knowledge point in each question according to the fuzzy attribute weight of each question for each knowledge point and the mastery degree of each user for each knowledge point;

[0039] obtain the actual score rate of each user based on the mastery degree of each user for the required knowledge point in each question, the failure rate and the guessing rate of the user;

[0040] determine the comprehensive mastery degree of each user for each knowledge point in combination with the actual score rate of each user and the correct answer probability of the user for each question in the user-question matrix, and take the knowledge point with a comprehensive mastery degree less than a preset threshold as the user weak knowledge point.

[0041] Optionally, the retrieval result integration module is specifically configured to introduce a plurality of member search engines to perform resource retrieval on the recommended learning path, integrate retrieval results under each member search engine, and obtain initial retrieval results; remove invalid retrieval results from the initial retrieval results, sort the remaining retrieval results, and obtain recommended results.

[0042] Optionally, after removing invalid retrieval results from the initial retrieval results, the retrieval result integration module sorts the remaining retrieval results to obtain recommended results, including:

[0043] After removing invalid retrieval results from the initial retrieval results, the remaining retrieval results are sorted in combination with the accuracy, recall rate, and nDCG of each member search engine to obtain recommended results.

[0044] The education resource retrieval method based on cognitive diagnosis and learning path provided by the application determines a test question-knowledge point matrix and a user-test question matrix according to a knowledge point set, wherein the test question-knowledge point matrix includes a video-knowledge point matrix, a practice question-knowledge point matrix, and other resource-knowledge point matrix, and the user-test question matrix includes a user-video matrix, a user-practice question matrix, and a user-other resource matrix; the test question-knowledge point matrix and the user-test question matrix are input into a cognitive diagnosis model, the cognitive diagnosis model is used to diagnose weak knowledge points of a user from multiple dimensions of videos, practice questions, and other resources, and the accuracy of determining weak knowledge points of the user is improved. In the process of planning a learning path by using an AprioriAll algorithm, the weight value of the weak knowledge points of the user is improved, so that more attention is paid to the weak knowledge points of the user in the recommended learning path, and the accuracy of the education resource retrieval result is ensured. In the application, a plurality of member search engines are introduced to perform resource retrieval on the recommended learning path, and the shortcomings of low coverage and insufficient personalization in the traditional retrieval method are solved.

[0045] The application will be further described in detail below with reference to the accompanying drawings and the application. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 is a flowchart of an education resource retrieval method based on cognitive diagnosis and a learning path provided by an embodiment of the application;

[0047] Figure 2 is a scene diagram of a knowledge point set in a data structure provided by an embodiment of the application;

[0048] Figure 3 is a structure diagram of a cognitive diagnosis model provided by an embodiment of the application;

[0049] Figure 4is a flowchart of scanning a knowledge point set using an AprioriAll algorithm provided by an embodiment of the present application.

[0050] Figure 5 is a resource retrieval flowchart provided by an embodiment of the present application.

[0051] Figure 6 is an educational resource retrieval device based on cognitive diagnosis and learning paths provided by an embodiment of the present application. DETAILED DESCRIPTION

[0052] The present application will be further described in detail below in conjunction with specific embodiments, but the embodiments of the present application are not limited thereto.

[0053] In order to solve the problem that the existing cognitive diagnosis method cannot accurately obtain the weak knowledge points of learners, an educational resource retrieval method based on cognitive diagnosis and learning paths is provided by an embodiment of the present application, referring to Figure 1 , Figure 1 is a flowchart of an educational resource retrieval method based on cognitive diagnosis and learning paths provided by an embodiment of the present application, specifically comprising the following steps:

[0054] Step S101, determining a test question-knowledge point matrix and a user-test question matrix according to a knowledge point set; the test question-knowledge point matrix includes a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix; the user-test question matrix includes a user-video matrix, a user-practice question matrix and a user-other resource matrix; the test question-knowledge point matrix is used to represent the correlation degree of test questions and knowledge points; the user-test question matrix is used to represent the answering situation of users to test questions.

[0055] In the embodiment of the present application, determining the test question-knowledge point matrix and the user-test question matrix according to the knowledge point set includes extracting the test question-knowledge point matrix and the user-test question matrix from a preset data set according to the knowledge point set.

[0056] The preset data set provided by the embodiment of the present application is obtained based on a MOOCCubeX data set. The MOOCCubeX data set is an open data warehouse for researchers in the field of natural language processing, knowledge graph, data mining and other related researches of large-scale online education, which contains JSON data of the answering situation of users to practice questions, the video watching situation and the learner comments, wherein JSON is an open standard file format and data exchange format.

[0057] The extracting of the test question-knowledge point matrix and the user-test question matrix from the data set according to the knowledge point set is specifically extracting a video-knowledge point matrix, a practice question-knowledge point matrix, an other resource-knowledge point matrix, a user-video matrix, a user-practice question matrix and a user-other resource matrix from the data set according to the knowledge point set. The other resource can be a comment of a user under information related to a knowledge point.

[0058] The video-knowledge point matrix represents the proportion of the related information of the knowledge point in the video, that is, the association degree of the knowledge point and the video. For example, in the video-knowledge point matrix, the row represents the video, the column represents the knowledge point, and the value of the a-th row and the b-th column is 0.3. The proportion of the b-th knowledge point in the a-th video is 0.3.

[0059] The practice question-knowledge point matrix represents the proportion of the related information of the knowledge point in the practice question, that is, the association degree of the knowledge point and the practice question. For example, in the practice question-knowledge point matrix, the row represents the practice question, the column represents the knowledge point, and the value of the c-th row and the d-th column is 0.5. The proportion of the d-th knowledge point in the c-th practice question is 0.5.

[0060] The other resource-knowledge point matrix represents the proportion of the related information of the knowledge point in the other resource, that is, the association degree of the knowledge point and the other resource. For example, in the other resource-knowledge point matrix, the row represents the other resource, the column represents the knowledge point, and the value of the e-th row and the f-th column is 0.7. The proportion of the f-th knowledge point in the e-th other resource is 0.7.

[0061] In the embodiment of the application, the user-video matrix represents the viewing time of the user to the video; the user-practice question matrix represents the answering situation of the user to the practice question; and the user-other resource matrix represents the mastering situation of the user to the other resource.

[0062] Based on the video-knowledge point matrix and the user-video matrix, the learning degree of the user to the video can be represented by the viewing time of the user to the video, and finally the mastering degree of the user to the related knowledge point in the video is obtained.

[0063] Based on the practice question-knowledge point matrix and the user-practice question matrix, the mastering degree of the user to a knowledge point can be represented by the correct rate of the user to the related practice question of the knowledge point.

[0064] Based on the other resource-knowledge point matrix and the user-other resource matrix, the mastering degree of the user to a knowledge point can be represented by the active degree of the user to the related content of the knowledge point.

[0065] Referring to Figure 2 , Figure 2is a scene schematic diagram provided by an embodiment of the present application with a data structure as a knowledge point set. Taking the data structure as the knowledge point set as an example, the knowledge point set can specifically include a tree, a graph, a layer sequence traversal algorithm, a sorting algorithm, a sequential list, a linked list, a greedy algorithm, a backtracking algorithm, a depth search traversal, and a breadth search traversal, etc. Exercises e1, e2, e3, and e4 represent exercises 1, 2, 3, and 4. The answer condition is used to represent whether the user correctly answers the exercise. The symbol “√” represents that the user correctly answers the exercise, and the symbol “×” represents that the user does not correctly answer the exercise. Exercises v1, v2, v3, and v4 represent videos 1, 2, 3, and 4. The video playback is used to represent whether the user completes the video. The symbol “√” represents that the user completes the video, and the symbol “×” represents that the user does not complete the video. When the other resource is a comment, exercises o1, o2, o3, and o4 represent comment conditions 1, 2, 3, and 4. The comment condition is used to represent whether the user makes a comment. The symbol “√” represents that the user makes a comment, and the symbol “×” represents that the user does not make a comment.

[0066] In an embodiment of the present application, the Q matrix can be used to represent the question-knowledge point matrix, where qvk=1 represents that the question Jv contains the knowledge point Sk, and qvk=0 represents that the question Jv does not contain the knowledge point Sk. The V matrix represents the user-question matrix, where ηuv=1 represents that the user Pu answers the question Jv correctly, and ηuv=0 represents that the user Pu does not answer the question Jv correctly.

[0067] In step S102, the question-knowledge point matrix and the user-question matrix are input into the cognitive diagnosis model, so that the cognitive diagnosis model obtains the weak knowledge points of the users by judging the comprehensive mastery degrees of each user on each knowledge point. The weak knowledge points of the users are the knowledge points with a comprehensive mastery degree less than a preset threshold.

[0068] In an embodiment of the present application, by taking the question-knowledge point matrix and the user-question matrix of each user as input, the cognitive diagnosis model can quantize the mastery level of each user on each knowledge point into a continuous numerical value, and finally output the comprehensive mastery degree of each user on each knowledge point.

[0069] In an embodiment of the present application, referring to Figure 3 , Figure 3 is a structure schematic diagram of the cognitive diagnosis model provided by an embodiment of the present application. The question-knowledge point matrix and the user-question matrix are input into the cognitive diagnosis model, so that the cognitive diagnosis model obtains the weak knowledge points of the users by judging the comprehensive mastery degrees of each user on each knowledge point in the following manner:

[0070] a) Calculate the fuzzy attribute weight of the question on the knowledge point.

[0071] The fuzzy attribute weights of each test question to each knowledge point are obtained according to the test question-knowledge point matrix; the fuzzy attribute weights are used to represent the degree of relevance between each test question and each knowledge point.

[0072] In this embodiment of the present invention, the fuzzy attribute weight α of the jth question to the kth knowledge point jk It can be understood as the degree of relevance between the jth test question and the kth knowledge point. When the jth test question has a high degree of relevance to the kth knowledge point, the fuzzy attribute weight of the jth test question on the kth knowledge point is high. Conversely, when the jth test question has a low degree of relevance to the kth knowledge point, the fuzzy attribute weight of the jth test question on the kth knowledge point is low. Where j = 1, 2, ..., M, k = 1, 2, ..., K, M represents the total number of test questions, and K represents the total number of knowledge points.

[0073] Specifically, if α jk =1, indicating that the jth test question is completely related to the kth knowledge point, that is, the jth test question directly contains the kth knowledge point. jk =0, indicating that the jth test question is completely irrelevant to the kth knowledge point, that is, the jth test question does not contain the kth knowledge point at all. jk The value range can be between [0,1].

[0074] In an embodiment of the present invention, the fuzzy attribute weights of the test questions to the knowledge points are determined based on the positive correlation weights between each test question and each knowledge point, the negative correlation weights between each test question and each knowledge point, and the difficulty of the test questions.

[0075] Specifically, α jk According to the positive correlation weight a between the jth test question and the kth knowledge point jk , the negative correlation weight b between the jth test question and the kth knowledge point jk And the difficulty of the j-th question determines θ j .

[0076] Among them, the positive correlation weight a jk Used to indicate the importance of the kth knowledge point in the jth test question. Negative correlation weight b jk Used to represent the influence of interference factors or non-target attributes of the k-th knowledge point in the j-th test question.

[0077] b) Determine the user's mastery of the knowledge points required in the test questions.

[0078] According to the fuzzy attribute weight of each test question for each knowledge point and the mastery degree of each user on each knowledge point, the mastery degree of each user on the knowledge points required in each test question is obtained.

[0079] In the embodiment of the present invention, the degree to which the i-th user has mastered the knowledge points required in the j-th test question can be expressed as ηji Indicates. ji It is a broad ability assessment indicator, usually used to reflect whether the user has a sufficient grasp of the knowledge points required in the test questions. ji The larger it is, the more fully the i-th user has mastered the knowledge points required for the j-th test question.

[0080] η ji The specific calculation method is:

[0081]

[0082] Among them, η ji According to q ik and q jk OK, q ik Indicates the degree of mastery of the kth knowledge point by the i-th user, and its value is in the range of [0, 1]. jk Represents the fuzzy attribute weight of the j-th test question to the k-th knowledge point. i=1,2,…,N o orN s , N o Represents a set of knowledge points, N s Represents a subset of the knowledge point set.

[0083] c) Determine the user's actual answer to the test questions.

[0084] Based on each user's mastery of the knowledge points required in each test question, the user's error rate and guessing rate are introduced to obtain each user's actual score rate.

[0085] In the embodiment of the present invention, the actual answer result of the i-th user to the j-th test question can be expressed by the actual score rate R ji To express, where R ji The value of is 0 or 1.

[0086] In this embodiment of the present invention, the user's real-world answering behavior can be modeled by introducing their mastery of the required knowledge points, the error rate s, and the guess rate g. The error rate s represents the probability that a user knows the knowledge point but answers the question incorrectly, while the guess rate g represents the probability that a user does not know the knowledge point but answers the question correctly. By introducing the error rate s and guess rate g, the user's mastery of the knowledge points can be further restored, resulting in the user's actual answer to the question.

[0087] d) Determine the user's comprehensive mastery of each knowledge point by combining each user's actual score rate and the probability of the user's correct answer to each question in the user-question matrix.

[0088] In this embodiment of the present invention, the probability μ of the correct answer of the i-th user to the j-th question isij and R ji The comprehensive mastery degree of each user on each knowledge point can be determined.

[0089] The correct answer probability μ of the ith user to the jth test question ij is a probability value, which can be directly used to reflect the possibility of the ith user correctly answering the jth test question, and the calculation method of μ ij is as follows:

[0090]

[0091] wherein, if q jk = 1, the expression is indicates that the jth test question contains the kth knowledge point, so the mastery degree of the user to the kth knowledge point needs to be considered, if q jk = 0, indicates that the jth test question does not contain the kth knowledge point. α ik indicates the mastery degree of the ith user to the kth knowledge point. q jk indicates the possibility that the jth test question contains the kth knowledge point, and takes the value of 0 or 1. When the value is 1, the jth test question contains the kth knowledge point; otherwise, it does not contain.

[0092] In the embodiment of the application, the knowledge point with the comprehensive mastery degree less than the preset threshold value is regarded as the weak knowledge point of the user.

[0093] The specific preset threshold value can be set by the technical personnel according to the requirement, which is not limited herein.

[0094] In step S103, the AprioriAll algorithm is used to scan the knowledge point set, and in combination with the weak knowledge point of the user, the recommended learning path is generated in the learning path planning process by increasing the weight value of the weak knowledge point of the user.

[0095] In the embodiment of the application, the AprioriAll algorithm is an association rule mining algorithm, which can be used to find the frequent set item in the knowledge point set.

[0096] Referring to Figure 4 , Figure 4is a flowchart of scanning a knowledge point set using an AprioriAll algorithm provided by an embodiment of the present application, determining information of a learner, i.e., information of a user, and determining a learning goal using the AprioriAll algorithm. In the AprioriAll algorithm, a minimum support threshold min_support is set, which can be used to filter the support threshold of a frequent item set. The comprehensive mastery degree of each knowledge point obtained through a cognitive diagnosis model is used to traverse the knowledge point set using depth-first traversal based on weak knowledge points or breadth-first traversal based on weak knowledge points, generate a frequent 1-item set, scan the knowledge point set, calculate the occurrence frequency of each individual item, retain items with a support degree greater than or equal to min_support, form a frequent 1-item set, and then generate a candidate t-item set using the frequent (t-1) item set. If any subset of a candidate item set is not frequent, the candidate item set cannot be frequent either. The candidate item set with a support degree greater than or equal to min_support is retained to form a frequent t-item set. When generating the frequent item set, the learning content of the user's weak knowledge points obtained through the cognitive diagnosis model is given a higher weight to improve its priority in recommendation. Association rules are generated in the frequent item set, a minimum confidence threshold is set, and learning content combinations with strong association are filtered out. Special attention is paid to content related to weak knowledge points to ensure that the recommendation contains these user weak knowledge points. According to the user's weak knowledge points, learning content that can compensate for these user weak points is preferentially recommended, and a learning path is recommended to help the user improve related knowledge. The effectiveness of the recommendation is evaluated through user feedback and learning effect, and the support and confidence thresholds are adjusted as necessary to optimize the recommendation result.

[0097] Specifically, the calculation method of the support degree includes:

[0098]

[0099] wherein Support(x) represents the support degree of the item set x; the support degree represents the frequency of the occurrence of a certain item set x in the entire knowledge point data set. count(x) represents the number of knowledge point records containing the item set x; and N represents the total number of records in the knowledge point data set.

[0100]

[0101] wherein confidence(x→y) represents the probability of containing the item set y under the condition of containing the item set x, x→y represents an association rule; and support(x∪y) represents the support degree of simultaneously containing the item set x and the item set y.

[0102] Step S104, introducing a plurality of member search engines to search for resources for the recommended learning path, and integrating the search results under each member search engine to obtain a recommended result.

[0103] In the embodiment of the present application, in view of the lack of coverage accuracy of the traditional education search system, a plurality of member search engines are introduced to integrate the search results to obtain a recommended result. Since the present application is directed to the field of education, the member search engines should be selected in combination with multiple dimensions such as education text information and education videos.

[0104] In the embodiment of the present application, the test question-knowledge point matrix and the user-test question matrix are determined according to the knowledge point set, wherein the test question-knowledge point matrix includes a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix, and the user-test question matrix includes a user-video matrix, a user-practice question matrix and a user-other resource matrix; the test question-knowledge point matrix and the user-test question matrix are input into the cognitive diagnosis model, and the cognitive diagnosis model is used to diagnose the weak knowledge points of the user from multiple dimensions of videos, practice questions and other resources, thereby improving the accuracy of the judgment of the weak knowledge points of the user. In the process of learning path planning using the AprioriAll algorithm, the weight value of the weak knowledge points of the user is improved, so that more attention is paid to the weak knowledge points of the user in the recommended learning path, and the accuracy of the education resource search result is ensured. In the embodiment of the present application, a plurality of member search engines are introduced to search for resources for the recommended learning path, thereby solving the shortcomings of low coverage and insufficient personalization in the traditional search method.

[0105] In one implementation, the cognitive diagnosis model includes a VFuzzy-CDF model, a CFuzzy-CDF model and a Fuzzy-CDF model, and the test question-knowledge point matrix and the user-test question matrix are input into the cognitive diagnosis model to make the cognitive diagnosis model judge the comprehensive mastery degree of each user for each knowledge point to obtain the weak knowledge points of the user, including:

[0106] The video-knowledge point matrix and the user-video matrix are input into the VFuzzy-CDF model to make the VFuzzy-CDF model judge the comprehensive mastery degree of each user for each knowledge point to obtain the first weak knowledge points of the user;

[0107] The other resource-knowledge point matrix and the user-other resource matrix are input into the CFuzzy-CDF model to make the CFuzzy-CDF model judge the comprehensive mastery degree of each user for each knowledge point to obtain the second weak knowledge points of the user;

[0108] The problem-knowledge point matrix and the user-problem matrix are input into the Fuzzy-CDF model, so that the Fuzzy-CDF model obtains the third weak knowledge point of the user by judging the comprehensive mastery degree of each user on each knowledge point;

[0109] The first weak knowledge point of the user, the second weak knowledge point of the user and the third weak knowledge point of the user are fused to obtain the weak knowledge point of the user.

[0110] In the embodiment of the application, the VFuzzy-CDF and the CFuzzy-CDF are improved based on the Fuzzy-CDF, and the Fuzzy-CDF is a knowledge tracking model. The existing Fuzzy-CDF can only be used to judge the mastery degree of the user on the knowledge point according to the problem, and the judgment basis is obtained by extracting the user-problem matrix and the problem-knowledge point matrix in the upstream knowledge point set.

[0111] In the VFuzzy-CDF provided in the embodiment of the application, the extracted matrix is improved, the video-knowledge point matrix and the user-video matrix are used as the input matrix of the VFuzzy-CDF, and the judgment basis of the model is improved to the learning degree of the user on the video and the proportion of the related knowledge point in the video.

[0112] In the CFuzzy-CDF provided in the embodiment of the application, the consideration of other resources is introduced, and specifically, the other resources can be the comment activity degree of the user in the related comment area. The input matrix of the model is improved to the other resource-knowledge point matrix and the user-other resource matrix.

[0113] For example, in the VFuzzy-CDF model, there are U users, V videos and K knowledge points. The video-knowledge point matrix and the user-video matrix are extracted, the learning degree of the user on the video is represented by the viewing time of the user on the video, the proportion of the related knowledge point in the video is combined, the mastery degree of the user on the knowledge point in the video is finally obtained, and the failure rate and the guessing rate are introduced to represent the probability that the user masters the knowledge point but does not watch the video and the probability that the user does not master the knowledge point but watches the video, to simulate the situation of the user watching the video in the real state.

[0114] For example, in the CFuzzy-CDF model, there are U users, C comments and K knowledge points. The other resource-knowledge point matrix and the user-other resource matrix are extracted, the mastery degree of the user on the knowledge point is represented by the comment activity degree of the user on the knowledge point, and the failure rate and the guessing rate are also introduced to represent the situation of the user publishing comments in the real state. The other resources can refer to comments.

[0115] Specifically, the Fuzzy-CDF is used to obtain the first weak knowledge point of the user by judging the comprehensive ability matching degree of the user on the corresponding knowledge point in the exercise; the VFuzzy-CDF is used to obtain the second weak knowledge point of the user by judging the comprehensive ability matching degree of the user on the corresponding knowledge point in the video; and the CFuzzy-CDF is used to obtain the third weak knowledge point of the user by judging the comprehensive ability matching degree of the user on the corresponding knowledge point in the comment.

[0116] In the embodiment of the application, the weak knowledge point of the user can be obtained by fusing the first weak knowledge point of the user, the second weak knowledge point of the user and the third weak knowledge point of the user.

[0117] In an implementation manner, a plurality of member search engines are introduced to perform resource retrieval on the recommended learning path, and the retrieval results under each member search engine are integrated to obtain a recommended result, including:

[0118] The plurality of member search engines are introduced to perform resource retrieval on the recommended learning path, and the retrieval results under each member search engine are integrated to obtain a primary retrieval result.

[0119] After removing invalid retrieval results from the primary retrieval result, the remaining retrieval results are sorted to obtain a recommended result.

[0120] Referring to Figure 5 , Figure 5 is a resource retrieval process schematic diagram provided by the embodiment of the application. A query interface introduces a plurality of member search engines to perform resource retrieval on the recommended learning path, and integrates the retrieval results under each member search engine to obtain a primary retrieval result. Each member search engine is egine 1, egine 2, …, egine n.

[0121] Then, invalid retrieval results are removed from the search result set, i.e., the primary retrieval result, specifically including judging whether the URL (uniform resource location) is the same to realize a deduplication operation, judging the abstract and title content, and removing dead links by ping website judgment.

[0122] Specifically, invalid link checking and duplicate information deduplication are performed, according to statistics, when users browse the query results, they usually only look at the first 20-30 results, therefore, the feasibility of the link of each query result can be checked. By sending a query request to the link, if the response code returns 404 or 500, the link is determined to be an invalid link and is directly deleted. The initial retrieval results come from each member search engine, therefore, there are usually more duplicate results in the initial retrieval results, these duplicate results, due to being able to be retrieved by multiple underlying retrieval systems, on the other hand, also represent the importance of the record, however, if they are not processed, it will further lead to the user forming an "information maze" and increasing the burden of the user. Therefore, the retrieval system needs to judge the duplication of the query results, by comparing the URL, title and abstract of two query results, if the URLs of the two query results are completely same, it is a duplicate result, if the URLs are different and the document title and abstract are same, it is also marked as a duplicate result.

[0123] In an implementation manner, after removing the invalid retrieval results from the initial retrieval results, the remaining retrieval results are sorted to obtain the recommended results, including:

[0124] After removing the invalid retrieval results from the initial retrieval results, the remaining retrieval results are sorted in combination with the accuracy, recall rate and nDCG of each member search engine to obtain the recommended results.

[0125] In the embodiment of the application, the calculation manner of the accuracy is:

[0126]

[0127] The calculation manner of the recall rate is:

[0128]

[0129] In the embodiment of the application, the accuracy and the recall rate are two indexes commonly used in information retrieval research, the accuracy refers to the proportion of correct duplicate web pages in the duplicate web pages detected by the search engine, and the recall rate refers to the ratio of the duplicate web pages found by the search engine to the total number of all duplicate web pages.

[0130] The nDCG (Normalized Discounted Cumulative Gain) is a retrieval metric designed for search engine evaluation. The weak knowledge points and the preceding and subsequent knowledge points are introduced, in all the remaining retrieval results, usually the first P retrieval results are taken for recommendation, therefore, the nDCG of the first P retrieval results, that is, the calculation manner of the nDCG p

[0131] ​

[0132] wherein P represents the length of the remaining search result list, usually the first P search results; rel represents the relevance score of the pth search result; p represents the ranking position of the search result; IDCG P represents the ideal DCG value of the first P results, the DCG value obtained after arranging the results in descending order of relevance.

[0133] After sorting the remaining search results by combining the accuracy, recall rate and nDCG of each member search engine, the recommended results are obtained, the recommended results are displayed, and the user is returned.

[0134] Based on the same inventive concept, the embodiment of the present application also provides an education resource retrieval device based on cognitive diagnosis and learning path, referring to Figure 6 , Figure 6 The education resource retrieval device based on cognitive diagnosis and learning path provided by the embodiment of the present application comprises:

[0135] The matrix determination module 601 is used for determining a test question-knowledge point matrix and a user-test question matrix according to the knowledge point set; the test question-knowledge point matrix comprises a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix; the user-test question matrix comprises a user-video matrix, a user-practice question matrix and a user-other resource matrix; the test question-knowledge point matrix is used for representing the correlation degree of test questions and knowledge points; the user-test question matrix is used for representing the answering condition of the user to the test questions;

[0136] The user weak knowledge point determination module 602 is used for inputting the test question-knowledge point matrix and the user-test question matrix into a cognitive diagnosis model, so that the cognitive diagnosis model obtains the user weak knowledge point by judging the comprehensive mastery degree of each user to each knowledge point; the user weak knowledge point is a knowledge point with a comprehensive mastery degree less than a preset threshold value;

[0137] The recommended learning path generation module 603 is used for scanning the knowledge point set by using an AprioriAll algorithm, combining the user weak knowledge point, and generating a recommended learning path by increasing the weight value of the user weak knowledge point in the learning path planning process.

[0138] The retrieval result integration module 604 is used for introducing a plurality of member search engines to perform resource retrieval on the recommended learning path, and integrating the retrieval results under each member search engine to obtain recommended results.

[0139] In the embodiment of the present application, the test question-knowledge point matrix and the user-test question matrix are determined according to the knowledge point set, wherein the test question-knowledge point matrix includes a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix, and the user-test question matrix includes a user-video matrix, a user-practice question matrix and a user-other resource matrix; the test question-knowledge point matrix and the user-test question matrix are input into the cognitive diagnosis model, and the cognitive diagnosis model is used to diagnose the weak knowledge points of the user from multiple dimensions of video, practice question and other resources, thereby improving the accuracy of the judgment of the weak knowledge points of the user. In the process of learning path planning by using the AprioriAll algorithm, the weight value of the weak knowledge points of the user is improved, so that more attention is paid to the weak knowledge points of the user in the recommended learning path, and the accuracy of the educational resource retrieval result is ensured. In the embodiment of the present application, a plurality of member search engines are introduced to perform resource retrieval on the recommended learning path, and the shortcomings of low coverage and insufficient personalization in the traditional retrieval method are solved.

[0140] Optionally, the cognitive diagnosis model includes a VFuzzy-CDF model, a CFuzzy-CDF model and a Fuzzy-CDF model, and the user weak knowledge point determination module is specifically configured to input the video-knowledge point matrix and the user-video matrix into the VFuzzy-CDF model, so that the VFuzzy-CDF model obtains the first weak knowledge points of the user by judging the comprehensive mastery of each user on each knowledge point; input the other resource-knowledge point matrix and the user-other resource matrix into the CFuzzy-CDF model, so that the CFuzzy-CDF model obtains the second weak knowledge points of the user by judging the comprehensive mastery of each user on each knowledge point; input the practice question-knowledge point matrix and the user-practice question matrix into the Fuzzy-CDF model, so that the Fuzzy-CDF model obtains the third weak knowledge points of the user by judging the comprehensive mastery of each user on each knowledge point; and fuse the first weak knowledge points of the user, the second weak knowledge points of the user and the third weak knowledge points of the user to obtain the weak knowledge points of the user.

[0141] Optionally, the user weak knowledge point determination module inputs the test question-knowledge point matrix and the user-test question matrix into the cognitive diagnosis model, so that the cognitive diagnosis model obtains the weak knowledge points of the user by judging the comprehensive mastery of each user on each knowledge point in the following manner:

[0142] obtain the fuzzy attribute weight of each test question on each knowledge point according to the test question-knowledge point matrix; the fuzzy attribute weight is used to represent the correlation degree between each test question and each knowledge point;

[0143] According to the fuzzy attribute weight of each knowledge point of each test question and the mastering degree of each user to each knowledge point, the mastering degree of each user to the required knowledge point in each test question is obtained;

[0144] Based on the mastering degree of each user to the required knowledge point in each test question, the actual score rate of each user is obtained by introducing the error rate and the guessing rate of the user;

[0145] The comprehensive mastering degree of each user to each knowledge point is determined by combining the actual score rate of each user and the correct answer probability of the user to each test question in the user-test question matrix, and the knowledge point with a comprehensive mastering degree less than a preset threshold is taken as a weak knowledge point of the user.

[0146] Optionally, the search result integration module is specifically configured to introduce resource search of the recommended learning path by a plurality of member search engines, integrate the search results under each member search engine to obtain initial search results, remove invalid search results from the initial search results, and sort the remaining search results to obtain recommended results.

[0147] Optionally, the search result integration module removes invalid search results from the initial search results, sorts the remaining search results to obtain recommended results, and includes the following steps.

[0148] After removing the invalid search results from the initial search results, the remaining search results are sorted by combining the accuracy, recall rate and nDCG of each member search engine to obtain the recommended results.

[0149] It should be noted that the terms "first", "second", and the like are used to distinguish similar objects, and do not necessarily have to be used to describe a particular order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of devices and methods consistent with some aspects of the present application.

[0150] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" and the like means that the specific features or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in the present specification.

[0151] Although the present application has been described in connection with certain embodiments, persons skilled in the art will understand and appreciate that many modifications can be made to the described embodiments and these modifications can be made without departing from the scope of the application. In the description of the application, the word "comprising" does not exclude other components or steps not mentioned, and the word "a" or "an" does not exclude a plurality. The mere fact that measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. The reference signs in the claims should not be construed as limiting the scope of the application.

[0152] The above description is further detailed in connection with specific preferred embodiments of the application, and it is not to be construed that the specific implementation of the application is limited to these descriptions. For those skilled in the art, without departing from the concept of the application, a number of simple deductions or substitutions can be made, and all of these should be considered as falling within the scope of protection of the application.

Claims

1. An education resource search method based on cognitive diagnosis and learning path, characterized in that, The education resource retrieval method comprises the following steps: According to the knowledge point set, a question-knowledge point matrix and a user-question matrix are determined; the question-knowledge point matrix comprises a video-knowledge point matrix, a practice question-knowledge point matrix and other resource-knowledge point matrix; the user-question matrix comprises a user-video matrix, a user-practice question matrix and a user-other resource matrix; the question-knowledge point matrix is used to represent the correlation degree of a question and a knowledge point; the user-question matrix is used to represent the answering condition of a user to a question; The question-knowledge point matrix and the user-question matrix are input into a cognitive diagnosis model, so that the cognitive diagnosis model determines the comprehensive mastery degree of each user to each knowledge point, and obtains a user weak knowledge point; the user weak knowledge point is a knowledge point with a comprehensive mastery degree less than a preset threshold value; The knowledge point set is scanned by using an AprioriAll algorithm, and the user weak knowledge point is combined to generate a recommended learning path by increasing the weight value of the user weak knowledge point in a learning path planning process; A plurality of member search engines are introduced to retrieve resources of the recommended learning path, and the retrieval results under each member search engine are integrated to obtain a recommended result; The question-knowledge point matrix and the user-question matrix are input into a cognitive diagnosis model, so that the cognitive diagnosis model determines the comprehensive mastery degree of each user to each knowledge point in the following manner: According to the question-knowledge point matrix, the fuzzy attribute weight of each question to each knowledge point is obtained; the fuzzy attribute weight is used to represent the correlation degree of each question and each knowledge point; According to the fuzzy attribute weight of each question to each knowledge point and the mastery degree of each user to each knowledge point, the mastery degree of each user to the required knowledge point in each question is obtained; Based on the mastery degree of each user to the required knowledge point in each question, the failure rate and the guessing rate of the user are introduced to obtain the actual score rate of each user; The actual score rate of each user and the correct answer probability of the user to each question in the question-knowledge point matrix are combined to determine the comprehensive mastery degree of each user to each knowledge point, and the knowledge point with a comprehensive mastery degree less than a preset threshold value is taken as a user weak knowledge point; The cognitive diagnosis model comprises a VFuzzy-CDF model, a CFuzzy-CDF model and a Fuzzy-CDF model, and the cognitive diagnosis model determines the comprehensive mastery degree of each user to each knowledge point by judging the comprehensive mastery degree of each user to each knowledge point, and obtains a user weak knowledge point, which comprises the following steps: The video-knowledge point matrix and the user-video matrix are input into the VFuzzy-CDF model, so that the VFuzzy-CDF model determines the comprehensive mastery degree of each user to each knowledge point, and obtains a first user weak knowledge point; input the other resource-knowledge point matrix and the user-other resource matrix into the CFuzzy-CDF model, so that the CFuzzy-CDF model obtains the second weak knowledge point of the user by judging the comprehensive mastery degree of each user to each knowledge point; input the exercise-knowledge point matrix and the user-exercise matrix into the Fuzzy-CDF model, so that the Fuzzy-CDF model obtains the third weak knowledge point of the user by judging the comprehensive mastery degree of each user to each knowledge point; fuse the first weak knowledge point of the user, the second weak knowledge point of the user and the third weak knowledge point of the user to obtain the weak knowledge point of the user.

2. The educational resource retrieval method of claim 1, wherein, introduce a plurality of member search engines to search resources for the recommended learning path, and integrate the search results under each member search engine to obtain a recommended result, including: introduce a plurality of member search engines to search resources for the recommended learning path, and integrate the search results under each member search engine to obtain a recommended result; remove invalid search results from the initial search results, and sort the remaining search results to obtain a recommended result.

3. The educational resource retrieval method of claim 2, wherein, remove invalid search results from the initial search results, and sort the remaining search results to obtain a recommended result, including: remove invalid search results from the initial search results, and sort the remaining search results by combining the accuracy, recall rate and nDCG of each member search engine to obtain a recommended result.

4. An educational resource search device based on cognitive diagnosis and learning path, characterized by, The educational resource searching device includes: a matrix determination module configured to determine a test question-knowledge point matrix and a user-test question matrix according to a knowledge point set; the test question-knowledge point matrix includes a video-knowledge point matrix, an exercise-knowledge point matrix and an other resource-knowledge point matrix; the user-test question matrix includes a user-video matrix, a user-exercise matrix and a user-other resource matrix; the test question-knowledge point matrix is used to represent the correlation degree between test questions and knowledge points; and the user-test question matrix is used to represent the answering situation of a user to test questions; a user weak knowledge point determination module configured to input the test question-knowledge point matrix and the user-test question matrix into a cognitive diagnosis model, so that the cognitive diagnosis model obtains a user weak knowledge point by judging the comprehensive mastery degree of each user to each knowledge point; the user weak knowledge point is a knowledge point with a comprehensive mastery degree less than a preset threshold value; a recommended learning path generation module configured to use an AprioriAll algorithm to scan the knowledge point set, combine the user weak knowledge point, and generate a recommended learning path by increasing the weight value of the user weak knowledge point in a learning path planning process; a search result integration module configured to introduce a plurality of member search engines to search resources for the recommended learning path, and integrate the search results under each member search engine to obtain a recommended result; The user weak knowledge point determination module inputs the test question-knowledge point matrix and the user-test question matrix into a cognitive diagnosis model, so that the cognitive diagnosis model determines the comprehensive mastery degree of each user to each knowledge point in the following manner to obtain the user weak knowledge point: The fuzzy attribute weight of each test question to each knowledge point is obtained according to the test question-knowledge point matrix; the fuzzy attribute weight is used to represent the correlation degree of each test question to each knowledge point; The mastery degree of each user to the required knowledge point in each test question is obtained according to the fuzzy attribute weight of each test question to each knowledge point and the mastery degree of each user to each knowledge point; The actual score rate of each user is obtained based on the mastery degree of each user to the required knowledge point in each test question, the failure rate and the guessing rate of the user; The comprehensive mastery degree of each user to each knowledge point is determined by combining the actual score rate of each user and the correct answer probability of the user to each test question in the user-test question matrix, and the knowledge point with a comprehensive mastery degree less than a preset threshold is taken as the user weak knowledge point. The cognitive diagnosis model includes a VFuzzy-CDF model, a CFuzzy-CDF model and a Fuzzy-CDF model, and the user weak knowledge point determination module is specifically configured to input the video-knowledge point matrix and the user-video matrix into the VFuzzy-CDF model, so that the VFuzzy-CDF model determines the comprehensive mastery degree of each user to each knowledge point to obtain a first user weak knowledge point; input the other resource-knowledge point matrix and the user-other resource matrix into the CFuzzy-CDF model, so that the CFuzzy-CDF model determines the comprehensive mastery degree of each user to each knowledge point to obtain a second user weak knowledge point; input the exercise-knowledge point matrix and the user-exercise matrix into the Fuzzy-CDF model, so that the Fuzzy-CDF model determines the comprehensive mastery degree of each user to each knowledge point to obtain a third user weak knowledge point; and the first user weak knowledge point, the second user weak knowledge point and the third user weak knowledge point are fused to obtain the user weak knowledge point.

5. The educational resource retrieval apparatus according to claim 4, wherein The retrieval result integration module is specifically configured to introduce resource retrieval of the recommended learning path by a plurality of member search engines, integrate the retrieval results under each member search engine to obtain initial retrieval results, and sort the remaining retrieval results after removing invalid retrieval results from the initial retrieval results to obtain recommended results.

6. The educational resource retrieval apparatus according to claim 5, wherein The retrieval result integration module sorts the remaining retrieval results after removing invalid retrieval results from the initial retrieval results to obtain recommended results, including: After removing invalid retrieval results from the initial retrieval results, the remaining retrieval results are sorted by combining the accuracy, recall rate and nDCG of each member search engine to obtain recommended results.

Images