Learning assistance system

Abstract

The invention provides a learning assistance system. The system comprises a teacher end, a student end and a server, the teacher end is used for uploading a to-be-updated question to the server; the student terminal is used for uploading a review demand to the server; the server is used for updating a to-be-updated question to a corresponding question bank based on a preset question bank knowledge graph to obtain an updated question bank, screening a question corresponding to a review knowledge point category from the updated question bank based on the review knowledge point category in a review demand, and sending the question as a review exercise to the student side; the preset question bank knowledge graph is constructed through a preset knowledge tree, the preset knowledge tree is extracted from review textbook content, real-time dynamic updating of the question bank and accurate personalized screening of questions are achieved by integrating the teacher end, the student end and the server, synchronization of the question bank content and the latest teaching content is guaranteed, and the accuracy of the question bank is improved. The personalized review requirements of the students are met, and the review effect of the students is effectively improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a learning assistance system. Background Art

[0002] In the field of education, with the rapid development of information technology, the application of learning assistance systems has become increasingly common. Learning assistance systems aim to optimize the learning process and improve students' learning efficiency and effectiveness through technical means such as data analysis and artificial intelligence.

[0003] At present, most of the learning assistance systems in related technologies use a pre-set question bank to determine the difficulty level of knowledge points by statistically analyzing the frequency and error rate of questions, and recommend exercises to students accordingly, so that students can achieve the purpose of reviewing and consolidating knowledge through these exercises.

[0004] However, since the question bank is fixed and the determination of the difficulty level depends on big data statistics, the question bank used by related technologies cannot timely reflect the latest teaching content and review needs, resulting in the questions that students review being out of touch with the actual teaching content. When the system is screening questions, it is unable to provide students with targeted review exercises, affecting the review effect. Summary of the invention

[0005] The present disclosure provides a learning assistance system.

[0006] According to a first aspect of the present disclosure, a learning assistance system is provided, which includes: a teacher terminal, a student terminal and a server; the teacher terminal is used to upload questions to be updated to the server; the student terminal is used to upload review requirements to the server; the server is used to update the questions to be updated to the corresponding question bank based on a preset question bank knowledge graph to obtain an updated question bank, and based on the review knowledge point category in the review requirement, filter out questions corresponding to the review knowledge point category from the updated question bank, and send the questions as review exercises to the student terminal; the preset question bank knowledge graph is constructed by a preset knowledge tree, and the preset knowledge tree is extracted from the review textbook content.

[0007] In some embodiments of the present disclosure, the server is used to obtain question knowledge points from the annotations in the question text to be updated, and determine the level corresponding to the question knowledge points in the preset question bank knowledge graph based on the knowledge point hierarchical structure in the preset question bank knowledge graph, and obtain the difficulty level of the question knowledge points based on the mapping relationship between the level and the difficulty level, so as to update the question to be updated to a question bank with the same difficulty level to obtain an updated question bank.

[0008] In some embodiments of the present disclosure, the server is used to determine the review knowledge point categories and review difficulty levels corresponding to the review needs, and based on the review knowledge point categories and review difficulty levels, search for corresponding review knowledge points from a preset question bank knowledge graph, so as to filter out questions matching the review knowledge points from the updated question bank as review exercises and send them to the student end. The review knowledge point categories include basic knowledge points and important knowledge points.

[0009] In some embodiments of the present disclosure, the server is also used to perform semantic analysis on the question bank text, extract the question bank titles and related knowledge points of the question bank titles in a preset knowledge tree, analyze the correlation between the titles and the related knowledge points based on the knowledge point hierarchical structure of the preset knowledge tree, and construct a question bank knowledge graph based on the correlation.

[0010] In some embodiments of the present disclosure, the server is used to select first related knowledge points and second related knowledge points at different levels based on the knowledge point hierarchy structure and determine the degree of association between the question and the first related knowledge point and the second related knowledge point, so as to associate the question with the first related knowledge point or the second related knowledge point according to the comparison result of the degree of association with a preset association threshold to obtain a question bank knowledge graph.

[0011] In some embodiments of the present disclosure, the server is used to determine the degree of association between the question and the first associated knowledge point and the second associated knowledge point according to the following formula:

[0012] in, is the correlation between the i-th associated knowledge point and the j-th associated knowledge point, is the first set of questions corresponding to the i-th knowledge point, is the second set of questions corresponding to the jth knowledge point, is the number of questions in the first question set, is the number of questions in the second question set, is the number of topics in the intersection of the first topic set and the second topic set.

[0013] In some embodiments of the present disclosure, the server is also used to obtain review textbook content uploaded by the student or teacher, and to grade the textbook text knowledge points extracted from the review textbook content to obtain a basic knowledge tree, so as to filter the knowledge tree using the distribution of knowledge points in the basic knowledge tree.

[0014] In some embodiments of the present disclosure, the server is used to construct a first important knowledge tree and a second important knowledge tree based on a historical question bank and wrong question association features, and optimize the knowledge tree based on the basic knowledge tree, the first important knowledge tree and the second important knowledge tree to obtain an optimized knowledge tree.

[0015] In some embodiments of the present disclosure, the server is used to use a preset knowledge point screening model to screen qualified knowledge points from all knowledge points corresponding to the historical question bank to obtain a first important knowledge tree; use a hierarchical clustering algorithm and a correlation analysis algorithm to screen wrong question features under a target feature category from wrong question-related features, and obtain a second important knowledge tree based on all knowledge points corresponding to the wrong question features; match and verify the knowledge trees based on the basic knowledge tree, the first important knowledge tree, and the second important knowledge tree, and optimize the knowledge tree based on the matching verification results to obtain an optimized knowledge tree.

[0016] In some embodiments of the present disclosure, the server is used to perform a first matching verification on all knowledge points at each level in the knowledge tree in turn using the first important knowledge tree and the second important knowledge tree based on the same-level matching rule; if the first matching verification fails, a cross-matching verification is performed on each first important knowledge point in the first important knowledge tree and each second important knowledge point in the second important knowledge tree; if the cross-matching verification is successful, the cross-matching knowledge points that are successfully cross-matched are determined, and based on the same-level matching rule, a second matching verification is performed on all basic knowledge points at each level in the basic knowledge tree in turn using the cross-matching knowledge points; if the second matching verification is successful, the matching knowledge points corresponding to the second successful match are integrated into the knowledge tree.

[0017] The learning assistance system provided by the present invention includes a teacher terminal, a student terminal and a server; the teacher terminal is used to upload questions to be updated to the server; the student terminal is used to upload review requirements to the server; the server is used to update the questions to be updated to the corresponding question bank based on a preset question bank knowledge graph to obtain an updated question bank, and based on the review knowledge point category in the review requirement, screen out questions corresponding to the review knowledge point category from the updated question bank, and send the questions as review exercises to the student terminal; the preset question bank knowledge graph is constructed by a preset knowledge tree, and the preset knowledge tree is extracted from the review teaching material content. By integrating the teacher terminal, the student terminal and the server, real-time dynamic updating of the question bank and accurate personalized screening of questions are realized, which not only ensures the synchronization of the question bank content with the latest teaching content, but also meets the personalized review needs of students, and effectively improves the students' review effect.

[0018] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. Figure 1 A schematic diagram of a learning assistance system provided by an embodiment of the present disclosure; Figure 2 A flowchart of a learning assistance method provided by an embodiment of the present disclosure; Figure 3 A structural schematic diagram of a learning assistance device provided by an embodiment of the present disclosure; Figure 4 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0020] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0021] In the field of education, with the rapid development of information technology, the application of learning assistance systems has become increasingly common. Learning assistance systems aim to optimize the learning process and improve students' learning efficiency and effectiveness through technical means such as data analysis and artificial intelligence.

[0022] At present, most of the learning assistance systems in related technologies use a pre-set question bank to determine the difficulty level of knowledge points by statistically analyzing the frequency and error rate of questions, and recommend exercises to students accordingly, so that students can achieve the purpose of reviewing and consolidating knowledge through these exercises.

[0023] However, since the question bank is fixed and the determination of the difficulty level relies on big data statistics, the system has a certain lag in responding to changes in teaching content and examination trends, resulting in the questions that students review being out of touch with the actual teaching content. When the system is screening questions, it is unable to provide students with targeted review exercises, affecting the review effect.

[0024] In order to solve the problems in the related technology, the learning assistance system proposed in the present disclosure realizes the dynamic update of the question bank and the accurate screening of questions by introducing the teacher side, the student side and the server, as well as the preset question bank knowledge graph. The teacher side can upload the questions to be updated in real time to ensure that the content of the question bank is consistent with the actual teaching content. At the same time, the student side can upload the review requirements, and the server will select the questions corresponding to the review knowledge point category from the updated question bank based on the preset question bank knowledge graph, and provide students with personalized review exercises. Not only the updating efficiency of the question bank is improved, but also the accuracy of question screening is enhanced, thereby effectively improving the review effect of students.

[0025] The following describes a learning assistance system, a learning assistance method, an apparatus, an electronic device, and a storage medium according to embodiments of the present disclosure with reference to the accompanying drawings.

[0026] Figure 1 Schematic diagram of a learning assistance system provided by an embodiment of the present disclosure. Figure 1 As shown, the system includes: Teacher terminal 1, student terminal 2 and server 3; The teacher terminal 1 is used to upload the questions to be updated to the server 3; The student terminal 2 is used to upload the review requirements to the server 3; Server 3 is used to update the questions to be updated to the corresponding question bank based on the preset question bank knowledge graph to obtain the updated question bank, and based on the review knowledge point category in the review requirements, screen out questions corresponding to the review knowledge point category from the updated question bank, and send the questions as review exercises to the student end; the preset question bank knowledge graph is constructed through a preset knowledge tree, and the preset knowledge tree is extracted from the review textbook content.

[0027] The present invention provides students with an efficient, personalized and interactive learning environment through close collaboration among the teacher, student and server, as well as the intelligent application of the preset question bank knowledge graph.

[0028] The teacher end is a user-friendly platform that allows teachers to upload new or revised questions to be updated to the server. The questions to be updated may be designed based on the latest teaching content, examination trends or student feedback, which is not limited in the embodiments of the present disclosure. The disclosed teacher uploads the questions to be updated in real time through the teacher end, ensuring the timeliness and accuracy of the question bank content, so that the students' learning materials are always consistent with the current teaching requirements.

[0029] The student terminal is the main interface for students to interact with the system. Students can upload their review requirements through the student terminal interface. The review requirements may include the knowledge points they wish to review, the desired difficulty of practice, or the question type preferences, etc. Students of the present disclosure can customize their review plans according to their actual situation and learning goals through the student terminal, thereby improving their learning efficiency.

[0030] The server is the core of the entire learning assistance system. It is responsible for receiving the updated questions uploaded by the teacher, and updating these questions to the corresponding question bank based on the preset question bank knowledge graph to form an updated question bank. At the same time, the server is also responsible for receiving the review requirements uploaded by the student, and accurately screening the questions based on these review requirements and the questions in the updated question bank, and sending the screened questions to the student as review exercises.

[0031] The present disclosure reduces the learning burden of students through the intelligent screening and personalized recommendation functions of the server, allowing students to focus on the knowledge points that really need to be reviewed. In addition, through the application of the preset question bank knowledge graph, the server can ensure that the screened questions are highly matched with the students' learning needs, thereby improving the review effect. The preset question bank knowledge graph enables the server to accurately understand the meaning, location and relationship of each knowledge point with other knowledge points, so that more intelligent and accurate decisions can be made when screening and recommending questions.

[0032] In some embodiments, the server is used to obtain question knowledge points from the annotations in the question text to be updated, and determine the level corresponding to the question knowledge points in the preset question bank knowledge graph based on the knowledge point hierarchical structure in the preset question bank knowledge graph, and obtain the difficulty level of the question knowledge points based on the mapping relationship between the level and the difficulty level, so as to update the question to be updated to a question bank with the same difficulty level to obtain an updated question bank.

[0033] In the learning assistance system, the server is not only responsible for data storage and transmission, but also undertakes the core tasks of knowledge processing and intelligent recommendation.

[0034] When the teacher uploads the questions to be updated to the server, the server will first parse the question texts. That is, through natural language processing (NLP) technology (such as entity recognition, keyword extraction, etc.), the server will identify and mark the knowledge points involved in the questions (i.e., question knowledge points), and then match the question knowledge points with the knowledge points in the preset question bank knowledge graph to ensure the consistency of the knowledge points.

[0035] The knowledge points in the preset question bank knowledge graph have a hierarchical structure (such as a tree structure), and each knowledge point has its corresponding level (such as first-level knowledge points, second-level knowledge points, etc.). The server will search for the corresponding level in the knowledge graph based on the extracted question knowledge points. There is a mapping relationship between the level of knowledge points and the difficulty level. For example, first-level knowledge points correspond to elementary difficulty, second-level knowledge points correspond to intermediate difficulty, and so on. The server determines the corresponding difficulty level based on the level of the question knowledge points. Based on the difficulty level of the question knowledge points, the server updates the questions to the question bank with the same difficulty level. The updated question bank will contain new questions, and the difficulty level of the questions will be consistent with the difficulty level of the question bank.

[0036] Specifically, in an optional embodiment of the present disclosure, assume that the title text of the question to be updated is: "Please explain Newton's first law." The server can extract the title knowledge point "Newton's first law" from the question to be updated, and search it in the preset question bank knowledge map to determine that "Newton's first law" belongs to the "Newton's law of motion" subcategory under "mechanics" and the level is a secondary knowledge point, so as to determine that the secondary knowledge point corresponds to the intermediate difficulty according to the mapping relationship. At this time, the server updates the question to be updated "Please explain Newton's first law." to the question bank of intermediate difficulty to ensure that the question difficulty distribution of the question bank is reasonable, so that students can choose appropriate questions to practice according to their own level.

[0037] In some embodiments, for students' review scenarios, the server can also determine the review knowledge point categories and review difficulty levels corresponding to the review needs, and based on the review knowledge point categories and review difficulty levels, search for corresponding review knowledge points from a preset question bank knowledge graph, so as to filter out questions matching the review knowledge points from the updated question bank as review exercises and send them to the student end. The review knowledge point categories include basic knowledge points and important knowledge points.

[0038] In the learning assistance system disclosed in the present invention, the server receives the review needs of the students, which may include the review scope (such as a certain subject, chapter) and the review goals (such as consolidating the basics, increasing the difficulty). The server parses the review needs through natural language processing (NLP) or predefined rules to determine the categories of knowledge points that need to be reviewed (basic knowledge points or important knowledge points) and the review difficulty level (such as elementary, intermediate, and advanced). Based on the review knowledge point categories and review difficulty levels obtained through the analysis, the server searches for the corresponding review knowledge points from the preset question bank knowledge graph, and selects questions that match the review knowledge points from the updated question bank based on the knowledge point categories and difficulty levels to which the questions belong, and sends the selected review questions to the student end for the students to practice.

[0039] In an optional embodiment of the present disclosure, it is assumed that the student's review requirement is: "Review the basic knowledge points of the mechanics part, the difficulty is elementary." At this time, the server can use natural language processing (NLP) to parse the review requirements to obtain the review scope of the mechanics part, the review knowledge point category is the basic knowledge points, and the review difficulty level is elementary, so as to find the basic knowledge points of the "mechanics" part from the knowledge graph according to the information obtained by the analysis, such as "Newton's First Law", "Composition and Decomposition of Forces", etc., and further confirm whether the difficulty level of these basic knowledge points is elementary. If the difficulty level is determined to be elementary, elementary difficulty questions that match knowledge points such as "Newton's First Law", "Composition and Decomposition of Forces", etc. are selected from the updated question bank, and elementary difficulty questions that match knowledge points such as "Newton's First Law", "Composition and Decomposition of Forces", etc. are sent to the student end as review exercises. Students can conduct targeted reviews through these questions to improve learning effects.

[0040] In some embodiments, the server is also used to perform semantic analysis on the question bank text, extract the question bank titles and associated knowledge points of the question bank titles in a preset knowledge tree, analyze the correlation between the titles and the associated knowledge points based on the knowledge point hierarchical structure of the preset knowledge tree, and construct a question bank knowledge graph based on the correlation.

[0041] Specifically, the server can use the BERT model to perform semantic analysis on the question bank text, extract key information from the question, and identify the question knowledge points involved in the question through natural language processing (NLP) technology. Based on the preset knowledge tree, the server matches the extracted question knowledge points with the nodes in the knowledge tree to determine the associated knowledge points of the question in the knowledge tree.

[0042] For example, the question "Please explain Newton's first law" will be associated with "Mechanics", "Newton's Laws of Motion", and "Newton's First Law" in the preset knowledge tree.

[0043] Afterwards, based on the hierarchical structure of the preset knowledge tree, the levels of the associated knowledge points that the questions may be associated with are further determined, so as to construct a question bank knowledge graph according to the degree of association between every two associated knowledge points at different levels.

[0044] Specifically, the server selects the first related knowledge point and the second related knowledge point at different levels based on the knowledge point hierarchy structure and determines the correlation between the question and the first related knowledge point and the second related knowledge point, so as to associate the question with the first related knowledge point or the second related knowledge point according to the comparison result of the correlation with the preset correlation threshold to obtain a question bank knowledge graph.

[0045] The server selects two related knowledge points that are most relevant to the question from the knowledge tree, that is, the second related knowledge point is a main knowledge point directly related to the question; and the first related knowledge point is a secondary knowledge point indirectly related to the question.

[0046] For example, for the question "Please explain Newton's first law", the related knowledge points extracted are first-level knowledge points (such as "mechanics"), second-level knowledge points (such as "Newton's laws of motion"), and third-level knowledge points (such as "Newton's first law"). The second related knowledge point is Newton's first law (third-level knowledge point); the first related knowledge point: Newton's laws of motion (second-level knowledge point).

[0047] In some implementations, the server may determine the degree of association between the topic and the first associated knowledge point and the second associated knowledge point according to the following formula:

[0048] in, is the correlation between the i-th associated knowledge point and the j-th associated knowledge point, is the first set of questions corresponding to the i-th knowledge point, is the second set of questions corresponding to the jth knowledge point, is the number of questions in the first question set, is the number of questions in the second question set, is the number of questions in the intersection of the first question set and the second question set. Wherein, i<j, that is, the i-th associated knowledge point is the first associated knowledge point, and the j-th associated knowledge point is the second associated knowledge point.

[0049] After calculating the correlation between the first related knowledge point and the second related knowledge point through the above formula 1, the server of the present invention can further compare the calculated correlation with a preset correlation threshold. If the correlation is greater than or equal to the preset correlation threshold, the question is associated with the second related knowledge point; if the correlation is less than the preset correlation threshold, the question is associated with the second related knowledge point.

[0050] This is because the knowledge points in the present disclosure are divided into hierarchical structures. When the correlation between a question and a certain knowledge point is high (greater than or equal to the preset correlation threshold), it means that the question is directly related to the knowledge point, so it is associated with the first associated knowledge point (usually a more specific and more directly related knowledge point); conversely, when the correlation is low (less than the preset correlation threshold), it means that the direct correlation between the question and the knowledge point is weak, so it is associated with the second associated knowledge point (usually a more general and indirectly related knowledge point). This logic ensures that the correlation between the question and the knowledge point is more accurate and reasonable, while making full use of the characteristics of the hierarchical structure of the knowledge point and optimizing the construction and management of the question bank knowledge graph.

[0051] The present disclosure calculates the relevance of each question in the above manner to associate each question with its corresponding knowledge point, thereby constructing a question bank knowledge graph. The nodes of the question bank knowledge graph include questions and knowledge points, and the edges represent the association between questions and knowledge points. The weight of the edges is determined by the relevance.

[0052] In some embodiments, since the question bank knowledge graph is constructed based on a preset knowledge tree, the server also needs to obtain the review textbook content uploaded by the student or teacher, and grade the textbook text knowledge points extracted from the review textbook content to obtain a basic knowledge tree, so as to use the distribution of knowledge points in the basic knowledge tree to filter and obtain the knowledge tree.

[0053] Specifically, the server can use text analysis technologies such as text mining, natural language, and semantic analysis to extract textbook knowledge points from the review textbook text of the review textbook content, and grade the textbook knowledge points (for example, grade them according to chapters, difficulty, learning stages, etc.), encode the identification of each textbook knowledge point, and obtain the identification coding value corresponding to each textbook knowledge point, so as to construct a set of identification vectors corresponding to the textbook knowledge point identification based on the identification coding value, generate a textbook knowledge point network U, and organize all textbook knowledge point networks into a basic knowledge tree Uori=[An, Bm, Ck].

[0054] In the present disclosure, the hierarchical identification can be divided into n levels. For example, when n=3, the textbook knowledge point network U=[An, Bm, Ck], the general expression is: U=(k, β), k is the knowledge point set, β is the edge set.

[0055] When n=3, the textbook knowledge point network U in the present disclosure can be a three-level classification system: the first-level classification An, the second-level classification Bm and the third-level classification Ck. Each level of classification represents a level in the textbook knowledge point network, where the first-level classification is the main node, the second-level classification is the slave node of the first-level classification, and the third-level classification is the slave node of the second-level classification. That is, the first-level classification An is the main node, and it is classified by chapter, and An=[A1, A2...An] can be obtained; the second-level classification Bm is the slave node of An, and it is classified by section, and Bm=[A1b1, A1b2, A2b1..Anbm] can be obtained; the third-level classification Ck is the slave node of Bm, and it is classified by unit results, and Ck=[A1b1c1, A1b2c2, A1b2c3..Anbmck] can be obtained.

[0056] Because textbook knowledge points can appear in the form of a single unit or a combination of multiple units depending on the difficulty and learning depth, when constructing the textbook knowledge point network, it is determined that the first-level classification exists in the form of a single-unit identification vector, and the second and third levels exist in the form of multiple units, with identification vectors superimposed in sequence.

[0057] The present disclosure forms a final basic knowledge tree by combining multiple textbook knowledge point networks corresponding to the identification code values ​​of textbook knowledge points. The basic knowledge tree disclosed in the present disclosure can cover all possible combinations of textbook knowledge points, providing a rich material library for the push of questions. In practical applications, the server can also further intelligently select textbook knowledge points of different difficulties for push according to the students' learning situation and needs. For example, for beginners or students who need to consolidate the basics, the system can push knowledge points of the first level classification; for students who want to study in depth, it can push knowledge points of the second or third level classification. This personalized push method not only improves learning efficiency, but also enhances students' interest and motivation in learning.

[0058] After obtaining the basic knowledge tree, the server in the present disclosure may count the number of basic knowledge points at each level in the basic knowledge tree, and determine the basic knowledge tree score based on the number of basic knowledge points at each level according to the following formula, so as to screen and obtain a knowledge tree according to the basic knowledge tree score result, that is, the preset knowledge tree of the present disclosure;

[0059] in, is the scoring result of the basic knowledge tree; Z is the number of basic knowledge points at the first level in the basic knowledge tree; X is the number of basic knowledge points at the second level in the basic knowledge tree; V is the number of basic knowledge points at the third level in the basic knowledge tree; M is the number of basic knowledge points at the nth level in the basic knowledge tree; is the weight of the number of basic knowledge points at the first level, is the weight of the number of basic knowledge points at the second level, is the weight of the number of basic knowledge points at the third level; is the weight of the number of basic knowledge points at the nth level. < < , because as the knowledge points become more refined, their importance accounts for a larger proportion.

[0060] The server of the present disclosure can compare the basic knowledge tree scoring result Q with the preset screening threshold T1. When Q is greater than or equal to T1, the knowledge point identification vector in the screened basic knowledge tree is used as the preset knowledge tree in the present disclosure. The preset screening threshold can be set by self-defined, average method, median, etc., which is not limited in the embodiments of the present disclosure.

[0061] In some embodiments, the server in the present disclosure may also optimize the preset knowledge tree obtained above, construct a first important knowledge tree and a second important knowledge tree based on the historical question bank and the associated features of wrong questions, and optimize the knowledge tree obtained above based on the basic knowledge tree, the first important knowledge tree and the second important knowledge tree to obtain an optimized knowledge tree, which is the preset knowledge tree of the present disclosure.

[0062] Among them, the server in the present disclosure can use a preset knowledge point screening model to screen qualified knowledge points from all knowledge points corresponding to the historical question bank to obtain a first important knowledge tree; use a hierarchical clustering algorithm and a correlation analysis algorithm to screen wrong question features under the target feature category from wrong question related features, and obtain a second important knowledge tree based on all knowledge points corresponding to the wrong question features; match and verify the knowledge tree based on the basic knowledge tree, the first important knowledge tree and the second important knowledge tree, and optimize the knowledge tree based on the matching verification results to obtain an optimized knowledge tree.

[0063] Specifically, the process of obtaining the first important knowledge tree is as follows: The server performs hierarchical processing on all knowledge points extracted from the history question bank text to obtain a history question bank knowledge tree, and determines the knowledge point score of each knowledge point in the history question bank knowledge tree according to a preset knowledge point screening model of the following formula, so as to screen qualified knowledge points from the knowledge point scores of each knowledge point to obtain a first important knowledge tree;

[0064] in, Score the knowledge point P in the history question bank knowledge tree, is the number of occurrences of knowledge point P, is the number of errors in knowledge point P, is the number of knowledge points at the level corresponding to knowledge point P, is the weight corresponding to the number of occurrences, The weight corresponding to the number of errors, is the weight corresponding to the number of knowledge points.

[0065] After obtaining the knowledge point score of each knowledge point, the present disclosure can compare the knowledge point score of each knowledge point. and the preset scoring threshold T2, when When it is greater than or equal to T2, the important knowledge point combinations are screened out to form the first important knowledge tree Znew.

[0066] The process of obtaining the second most important knowledge tree is as follows: The server disclosed in the present invention can obtain the wrong question correlation characteristics of the subjective and objective factors that affect students' wrong answers collected by the teaching management system, and use cluster analysis and correlation evaluation models to screen out the wrong question characteristics caused by objective factors, and then determine the feature group or features with the greatest impact, and accordingly construct the second important knowledge tree Onew that focuses on students' mastery of knowledge points.

[0067] The present disclosure collects factors that affect wrong answers through a teaching management system, including emotions, concentration, answering style, knowledge point mastery, answering efficiency, question difficulty level, etc., and simulates test scenarios through the teaching management system to collect students' answering video images, and can obtain subjective factor feature data sets through Single Shot Multibox Detector (SSD) facial expression recognition technology and eye tracking technology; obtain students' answering situations collected by computers through the teaching management system to obtain objective factor feature data sets. After obtaining these wrong answer-related features, the teaching management system in the present disclosure can feed back the obtained wrong answer-related features to the server.

[0068] Among them, subjective factors include: emotional characteristics (such as positive, indifferent, negative, etc.), concentration characteristics (such as concentration, distraction, susceptibility to external factors, etc.), test-taking style characteristics (such as carefulness, carelessness, lack of patience, etc.). Objective factors include: knowledge point mastery characteristics (such as proficiency, proficiency, unproficiency, etc.), test-taking efficiency characteristics (such as fast, medium, slow, etc.), and question difficulty level characteristics (such as simple, ordinary, difficult, etc.).

[0069] The server disclosed in the present invention can use clustered wrong question related feature groups to analyze the influence of different groups of wrong question related features on wrong questions. That is, the present invention can use random combination of all wrong question related features to find out the objective factors that really affect wrong questions. This is because there is data interference from subjective factors, and it is impossible to accurately judge whether students have mastered the corresponding knowledge points. Therefore, the present invention needs to filter out the wrong question feature data set caused by objective factors from the wrong question related feature data set, thereby improving the accuracy of the second important knowledge tree.

[0070] Among them, when performing cluster analysis, the present disclosure needs to first perform data preprocessing on the acquired wrong question related features, such as processing missing values ​​and outliers for the wrong question related features to ensure data quality, and perform Z-score standardization on the wrong question related features.

[0071] Afterwards, due to the difficulty of the questions, the efficiency of answering the questions is slow, which is a normal phenomenon. Therefore, the server disclosed in the present invention needs to first use a hierarchical clustering algorithm to group the pre-processed wrong question related features, and use the Pearson correlation coefficient distance algorithm of the following formula to calculate the distance matrix between each wrong question related feature to obtain the first correlation analysis result.

[0072]

[0073] in, is the result of the first correlation analysis, , is the average value of wrong question related features a and wrong question related features b, where a and b are wrong question related features. The value range is [0, 1]. The closer the value is to 1, the looser the linear correlation between the wrong question association feature and the wrong question feature.

[0074] Based on the above first correlation analysis results, hierarchical clustering is selected to gradually merge or split clusters to generate a dendrogram to obtain the correlation distribution between the associated features of wrong questions.

[0075] For example, the present disclosure may select the shortest distance linking method: d(Ccluster,Dcluster)=min{d(c,d):c∈C,d∈D}, where d(Ccluster,Dcluster) represents the distance between cluster C and cluster D, and d(Ccluster,Dcluster) represents the distance between a point c in cluster C and a point d in cluster D. Determine the number of groups k of wrong question-related features, and assign the wrong question-related features to the corresponding groups to obtain the correlation distribution between the wrong question-related features.

[0076] Afterwards, the server analyzes each set of wrong question-related features, evaluates their relationship with the wrong question features again, and obtains the second correlation analysis result (i.e., the correlation distribution between the wrong question-related feature sets is generated based on the second correlation analysis result, to determine whether there is correlation between the wrong question-related feature sets) through the pre-trained correlation evaluation model: The relevance evaluation model is as follows:

[0077] Among them, y is the result of the second correlation analysis, is the intercept, is the partial regression coefficient, and x1, x2...xn are the associated features of wrong questions.

[0078] The present disclosure selects the wrong question-associated feature group or features that have the greatest impact on wrong questions through the first correlation analysis result and the second correlation analysis result (i.e., selects the wrong question-associated feature with the maximum mean value of the first correlation analysis result and the second correlation analysis result of each wrong question-associated feature, or selects the wrong question-associated feature with the maximum value of the weighted fusion of the first correlation analysis and the second correlation analysis result of each wrong question-associated feature), and determines the feature category of each screened wrong question-associated feature (feature categories include subjective categories and objective categories). The present disclosure uses the wrong question-associated feature whose feature category is the target feature category (i.e., objective category) as the wrong question feature, thereby constructing the second important knowledge tree Onew.

[0079] Among them, the subjective category reflects the students' own psychology, not the reason for making mistakes due to unfamiliarity with knowledge points. What needs to be improved is their own psychological quality; the objective category reflects that students make mistakes due to insufficient knowledge points. Because the subjective category requires students to adjust themselves or a third party to assist in guidance, and only the objective category is the essential reason for making mistakes due to insufficient knowledge points, so the objective category is used as the target feature category.

[0080] After obtaining the first important knowledge tree and the second important knowledge tree through the above content, the present disclosure can match and verify the knowledge tree obtained at the beginning based on the basic knowledge tree, the first important knowledge tree and the second important knowledge tree, and optimize the knowledge tree based on the matching verification result to obtain an optimized knowledge tree.

[0081] Specifically, based on the same-level matching rule, the server uses the first important knowledge tree and the second important knowledge tree to perform a first matching verification on all knowledge points at each level in the knowledge tree in turn; if the first matching verification fails, a cross-matching verification is performed on each first important knowledge point in the first important knowledge tree and each second important knowledge point in the second important knowledge tree; if the cross-matching verification is successful, the cross-matching knowledge points that are successfully cross-matched are determined, and based on the same-level matching rule, the server uses the cross-matching knowledge points to perform a second matching verification on all basic knowledge points at each level in the basic knowledge tree in turn; if the second matching verification is successful, the matching knowledge points corresponding to the second successful match are integrated into the knowledge tree.

[0082] It should be noted that, for the matching verification process in the present disclosure, if the matching verification of the current level fails, the matching verification of the next level will be automatically performed until all current levels have been matched and different operations are performed according to the matching verification results. The cross-matching in the present disclosure refers to the matching verification of the same level between the first important knowledge tree and the second important knowledge tree after the first matching verification of the knowledge tree fails, to determine whether there are matching knowledge points in the first important knowledge tree and the second important knowledge tree, so as to avoid the situation where the knowledge point recognition is wrong due to system errors.

[0083] If the first matching verification is successful, the matching knowledge point that is successfully matched is bound to the question bank and imported into the knowledge tree. If the cross matching verification fails or the second matching verification fails, the knowledge tree currently undergoing matching verification is stored in the candidate library.

[0084] The server may obtain the first matching result and the second matching result according to the following formula to comprehensively obtain the matching verification result of each matching verification;

[0085] in, is the first matching result between the knowledge tree G and the knowledge tree H, where the knowledge tree G and the knowledge tree H are the basic knowledge tree, the first important knowledge tree, the second important knowledge tree, and any two of the knowledge trees that need to be matched and verified. Represents the number of knowledge points in the intersection of knowledge tree G and knowledge tree H, Represents the number of knowledge points in the union of knowledge tree G and knowledge tree H;

[0086] in, is the second matching result between knowledge tree G and knowledge tree H, n is the nth knowledge point, N is the total number of knowledge points after knowledge tree G and knowledge tree H are aligned, is the nth knowledge point vector in the knowledge tree G, is the mean value of the nth knowledge point vector after the knowledge tree G and the knowledge tree H are aligned, is the nth knowledge point vector in the knowledge tree H.

[0087] In the present disclosure, when the first match verification result is greater than or equal to the first match verification threshold, and the second match verification result is greater than the second match verification threshold, the current match verification result is determined to be a successful match.

[0088] It is understandable that, since the verification matching in the present disclosure adopts the principle of same-level matching, if only relying on the matching verification algorithm of the first formula (for example, the Jaccard similarity algorithm), a situation may be encountered: even if the matching degree of two objects seems similar overall (such as the output result of q (1, 0, 1)), there is a mismatch in the more detailed secondary dimension (such as the secondary dimension value is 0), which does not meet the consistency standard. Therefore, in order to make up for this deficiency, the present disclosure further combines the matching verification algorithm of the second formula (for example, the improved cosine similarity algorithm) for verification. The improved cosine similarity algorithm is improved compared with other algorithms in measuring the differences in the values ​​of each dimension, but its accuracy still has certain limitations. Therefore, by combining these two algorithms, the present disclosure can achieve complementary advantages, thereby improving the accuracy and reliability of matching.

[0089] In summary, the learning assistance system provided by the present invention includes a teacher end, a student end and a server; the teacher end is used to upload the questions to be updated to the server; the student end is used to upload the review requirements to the server; the server is used to update the questions to be updated to the corresponding question bank based on the preset question bank knowledge graph, obtain the updated question bank, and based on the review knowledge point category in the review requirements, screen out questions corresponding to the review knowledge point category from the updated question bank, and send the questions as review exercises to the student end; the preset question bank knowledge graph is constructed by a preset knowledge tree, and the preset knowledge tree is extracted from the review textbook content. By integrating the teacher end, the student end and the server, the real-time dynamic update of the question bank and the accurate personalized screening of the questions are realized, which not only ensures the synchronization of the question bank content with the latest teaching content, but also meets the personalized review needs of the students, and effectively improves the students' review effect.

[0090] based on Figure 1 The learning assistance system shown, Figure 2 The following is a flow chart of a learning assistance method provided by an embodiment of the present disclosure. Figure 2 As shown, the method is mainly applied to a server in a learning assistance system, and the method includes: Step 101: Based on the preset question bank knowledge graph, the questions to be updated received from the teacher end are updated to the corresponding question bank to obtain an updated question bank. The preset question bank knowledge graph is constructed through a preset knowledge tree, and the preset knowledge tree is extracted from the review textbook content.

[0091] Step 102: Based on the review knowledge point categories in the review requirements received from the student end, select questions corresponding to the review knowledge point categories from the updated question bank, and send the questions to the student end as review exercises, so that the student end can consolidate learning based on the review questions.

[0092] In some embodiments, the principles of the embodiments of step 101 and step 102 are the same as those described above. Figure 1 The principles of the related embodiments shown are the same, and can be referred to Figure 1 The relevant description of the illustrated embodiment will not be repeated here.

[0093] In summary, the technical solution provided by the present invention updates the questions to be updated received from the teacher end to the corresponding question bank based on the preset question bank knowledge graph to obtain an updated question bank, and the preset question bank knowledge graph is constructed by a preset knowledge tree, and the preset knowledge tree is extracted from the review textbook content; based on the review knowledge point category in the review requirements received from the student end, the questions corresponding to the review knowledge point category are screened out from the updated question bank, and the questions are sent to the student end as review exercises, so that the student end can consolidate its learning based on the review questions, realize real-time dynamic update of the question bank and accurate personalized screening of questions, which not only ensures the synchronization of the question bank content with the latest teaching content, but also meets the students' personalized review needs, not only improves the efficiency of question bank management, but also provides intelligent support for teaching and learning, and ultimately improves teaching quality and learning effects.

[0094] Corresponding to the above-mentioned learning assistance method, the present invention also provides a learning assistance device. Since the device embodiment of the present invention corresponds to the above-mentioned method embodiment, details not disclosed in the device embodiment can be referred to the above-mentioned method embodiment, and will not be repeated in the present invention.

[0095] Figure 3 A structural diagram of a learning assistance device provided by an embodiment of the present disclosure is shown in FIG. Figure 3 As shown, the device comprises: The question bank processing unit 310 is used to update the questions to be updated received from the teacher end to the corresponding question bank based on the preset question bank knowledge graph to obtain an updated question bank. The preset question bank knowledge graph is constructed through a preset knowledge tree, and the preset knowledge tree is extracted from the review textbook content.

[0096] The review auxiliary unit 320 is used to filter out questions corresponding to the review knowledge point categories from the updated question bank based on the review knowledge point categories in the review requirements received from the student end, and send the questions as review exercises to the student end, so that the student end can consolidate learning based on the review questions.

[0097] It should be noted that the above explanation of the method embodiment is also applicable to the device of this embodiment, and the principle is the same, which is not limited in this embodiment.

[0098] Based on the above Figure 2 The method shown in the figure, accordingly, this embodiment also provides a computer program product, including a computer program, which implements the above-mentioned Figure 2 The method shown.

[0099] Based on the above Figure 2 The method shown in the embodiment also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned Figure 2 The method shown.

[0100] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of the present application.

[0101] like Figure 4 The figure shows a hardware structure diagram of an electronic device of the present invention, including: at least one processor 401; and, A memory 402 is communicatively connected to at least one of the processors 401; wherein, The memory 402 stores instructions that can be executed by at least one of the processors. The instructions are executed by at least one of the processors to enable the at least one of the processors to perform the learning assistance method as described above.

[0102] Figure 4 A processor 401 is taken as an example.

[0103] The electronic device may further include: an input device 403 and a display device 404 .

[0104] The processor 401, the memory 402, the input device 403 and the display device 404 may be connected via a bus or other means, and the figure takes the connection via a bus as an example.

[0105] The memory 402 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the review content generation method in the embodiment of the present application, for example, Figure 2 The processor 401 executes various functional applications and data processing by running the non-volatile software programs, instructions and modules stored in the memory 402, that is, the learning assistance method in the above embodiment is implemented.

[0106] The memory 402 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created according to the use of the review content generation method, etc. In addition, the memory 402 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 402 may optionally include a memory remotely arranged relative to the processor 401, and these remote memories may be connected to the device for executing the review content generation method via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0107] The input device 403 can receive user clicks and generate signal inputs related to user settings and function controls of the review content generation method. The display device 404 can include display devices such as a display screen.

[0108] The one or more modules are stored in the memory 402 and, when executed by the one or more processors 401 , execute the learning assistance method in any of the above method embodiments.

[0109] Optionally, the above-mentioned physical device may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.

[0110] Those skilled in the art will appreciate that the above-mentioned physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different arrangements of components.

[0111] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the above-mentioned physical device, and supports the operation of the information processing program and other software and / or programs. The network communication module is used to realize the communication between the components inside the storage medium, and the communication with other hardware and software in the information processing physical device.

[0112] Through the description of the above implementation methods, the technicians in this field can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by hardware. By applying the scheme of this embodiment, compared with the current prior art, this embodiment is based on the preset question bank knowledge graph, updates the to-be-updated questions received from the teacher end to the corresponding question bank, and obtains the updated question bank. The preset question bank knowledge graph is constructed by a preset knowledge tree, and the preset knowledge tree is extracted from the review teaching material content; based on the review knowledge point category in the review demand received from the student end, select the questions corresponding to the review knowledge point category from the updated question bank, and send the questions as review exercises to the student end, so that the student end can consolidate learning based on the review questions, realize the real-time dynamic update of the question bank and the accurate personalized screening of the questions, which not only ensures the synchronization of the question bank content with the latest teaching content, but also meets the personalized review needs of students, not only improves the efficiency of question bank management, but also provides intelligent support for teaching and learning, and finally improves the teaching quality and learning effect.

[0113] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0114] The above is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features applied for herein.

Claims

1. A learning assistance system, characterized in that: The system comprises: Teacher side, student side and server; The teacher terminal is used to upload the topic to be updated to the server; The student terminal is used to upload the review requirements to the server; The server is used to update the to-be-updated questions to the corresponding question bank based on the preset question bank knowledge graph to obtain an updated question bank, and based on the review knowledge point category in the review requirement, screen out questions corresponding to the review knowledge point category from the updated question bank, and send the questions as review exercises to the student terminal; the preset question bank knowledge graph is constructed by a preset knowledge tree, and the preset knowledge tree is extracted from the review teaching material content.

2. The system according to claim 1, characterized in that The server is used to obtain question knowledge points from the annotations in the question text to be updated, and determine the level corresponding to the question knowledge points in the preset question bank knowledge graph based on the knowledge point hierarchical structure in the preset question bank knowledge graph, and obtain the difficulty level of the question knowledge points based on the mapping relationship between the level and the difficulty level, so as to update the question to be updated to a question bank with the same difficulty level to obtain an updated question bank.

3. The system according to claim 1, characterized in that The server is used to determine the review knowledge point category and review difficulty level corresponding to the review requirement, and based on the review knowledge point category and review difficulty level, search for the corresponding review knowledge point from the preset question bank knowledge graph, so as to filter out questions matching the review knowledge point from the updated question bank as the review exercises and send them to the student end, and the review knowledge point category includes basic knowledge points and important knowledge points.

4. The system according to claim 1, characterized in that The server is also used to perform semantic analysis on the question bank text, extract the question bank titles and the associated knowledge points of the question bank titles in the preset knowledge tree, analyze the correlation between the titles and the associated knowledge points based on the knowledge point hierarchical structure of the preset knowledge tree, and construct a question bank knowledge graph based on the correlation.

5. The system according to claim 4, characterized in that The server is used to select first associated knowledge points and second associated knowledge points at different levels based on the knowledge point hierarchical structure and determine the degree of association between the question and the first associated knowledge points and the second associated knowledge points, so as to associate the question with the first associated knowledge point or the second associated knowledge point according to the comparison result of the association degree with a preset association threshold to obtain a question bank knowledge graph.

6. The system according to claim 5, characterized in that The server is used to determine the degree of association between the question and the first associated knowledge point and the second associated knowledge point according to the following formula; in, is the correlation between the i-th associated knowledge point and the j-th associated knowledge point, is the first question set corresponding to the i-th knowledge point, is the second question set corresponding to the j-th knowledge point, is the number of topics in the first topic set, is the number of topics in the second topic set, is the number of topics in the intersection of the first topic set and the first topic set.

7. The system according to claim 1, characterized in that The server is also used to obtain the review teaching material content uploaded by the student terminal or the teacher terminal, and to perform hierarchical processing on the teaching material text knowledge points extracted from the review teaching material content to obtain a basic knowledge tree, so as to filter and obtain a knowledge tree using the distribution of knowledge points in the basic knowledge tree.

8. The system according to claim 7, characterized in that The server is used to construct a first important knowledge tree and a second important knowledge tree based on a historical question bank and wrong question association features, and optimize the knowledge tree based on the basic knowledge tree, the first important knowledge tree and the second important knowledge tree to obtain an optimized knowledge tree.

9. The system according to claim 8, characterized in that The server is used to use a preset knowledge point screening model to screen qualified knowledge points from all knowledge points corresponding to the historical question bank to obtain the first important knowledge tree; use a hierarchical clustering algorithm and a correlation analysis algorithm to screen wrong question features under a target feature category from wrong question related features, and obtain the second important knowledge tree based on all knowledge points corresponding to the wrong question features; The knowledge tree is matched and verified based on the basic knowledge tree, the first important knowledge tree and the second important knowledge tree, and the knowledge tree is optimized based on the matching and verification result to obtain the optimized knowledge tree.

10. The system according to claim 9, characterized in that The server is used to perform a first matching verification on all knowledge points at each level in the knowledge tree in sequence by using the first important knowledge tree and the second important knowledge tree based on a same-level matching rule; If the first matching verification fails, cross-matching verification is performed on each first important knowledge point in the first important knowledge tree and each second important knowledge point in the second important knowledge tree; if the cross-matching verification is successful, the cross-matching knowledge points that are successfully cross-matched are determined, and based on the same-level matching rule, the cross-matching knowledge points are used to perform a second matching verification on all basic knowledge points at each level in the basic knowledge tree in sequence; If the second matching verification is successful, the matching knowledge points corresponding to the second matching success are integrated into the knowledge tree.

Images