Intelligent Teaching Aided System and Method Based on Behavioral Data and Knowledge Graph

Abstract

The present invention discloses an intelligent teaching assistance system and method based on behavior data and knowledge graphs, which relates to the technical field of teaching assistance. It includes evaluating the mastery level of a first student over different learned knowledge points to obtain knowledge mastery data; obtaining defective knowledge points from the knowledge mastery data, obtaining the knowledge graph of the first student, and evaluating the influence degree of the knowledge points in the knowledge graph on the learning of the defective knowledge points by the first student; obtaining the historical learning behavior records and historical knowledge mastery data of other students, and combining the historical learning behavior records and defective learning data of the first student to analyze the learning reference value of other students to the first student to obtain target reference students; obtaining the historical learning data of the target reference students and combining the defective learning data of the first student to generate teaching reference data for the first student, and pushing the teaching reference data to the teacher in charge of the first student to assist the teacher in charge in teaching the first student.

Description

Technical Field

[0001] The present invention relates to the technical field of teaching aids, and specifically to an intelligent teaching aid system and method based on behavior data and knowledge graphs. Background Art

[0002] Behavior data refers to the data generated by students during the learning process. A knowledge graph is a structured way of representing knowledge, which organizes knowledge in the form of a graph. Therefore, with the continuous development of teaching intelligence, the requirements for teaching aid systems are getting higher and higher. The application of behavior data and knowledge graphs can just solve various problems arising in teaching aid systems. The application of behavior data and knowledge graphs in teaching aid systems includes, but is not limited to, the following benefits: 1. Personalized learning. By analyzing the behavior data of students, the intelligent teaching aid system can identify the learning characteristics of students. The knowledge graph is conducive to understanding the learning characteristics and needs of students, so as to dynamically adjust the learning content and achieve personalized learning; 2. Optimization of teaching strategies. Teachers can understand the learning progress of all students through behavior data and adjust teaching strategies in a timely manner. The use of the knowledge graph can provide teachers with a global perspective of teaching and help teachers optimize teaching strategies.

[0003] In traditional teaching aid systems, students can use Internet devices to independently learn various information related to courses, and teachers can also publish learning content through teaching platforms for students to study. However, there are deficiencies in this. First of all, the number of students taught by teachers is large, and the learning situations of each student are different. Even if two students make mistakes on the same question, the fact that the question is wrong is only the result and cannot indicate that their mastery of the knowledge points involved in the question is the same. Moreover, students with different levels have different absorption degrees of the same teaching content. This will not only cause teachers to be unable to master the learning situations of students and unable to formulate teaching plans, but also lead to the content learned by students not matching their learning abilities, resulting in insignificant improvement in students' grades after using the teaching aid system, and even leading to a decline in students' learning enthusiasm and reduced learning motivation. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent teaching aid system and method based on behavior data and knowledge graphs to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An intelligent teaching aid method based on behavior data and knowledge graphs, the method includes:

[0006] Step S100: Obtain the historical learning behavior records of the first student, extract historical learning behavior data from the historical learning behavior records, evaluate the mastery levels of the first student for different learned knowledge points, and obtain knowledge mastery data;

[0007] Step S200: Obtain the defective knowledge points from the knowledge mastery data, obtain the knowledge graph of the first student, evaluate the influence degree of the knowledge points in the knowledge graph on the learning of the defective knowledge points by the first student, and obtain defective learning data;

[0008] Step S300: Obtain the historical learning behavior records and historical knowledge mastery data of other students, and analyze the learning reference value of other students for the first student in combination with the historical learning behavior records and defective learning data of the first student, so as to obtain target reference students;

[0009] Step S400: Obtain the historical learning data of the target reference students, and generate teaching reference data for the first student in combination with the defective learning data of the first student, and push the teaching reference data to the teacher in charge of the first student to assist the teacher in teaching the first student.

[0010] Further, step S100 includes:

[0011] Step S101: Monitor and record the learning behaviors of the student on the teaching platform to obtain the historical learning behavior records of the first student, and extract historical learning behavior data from the historical learning behavior records. The historical learning behavior data includes several knowledge points involved by the first student in the learning behaviors;

[0012] Step S102: According to the different involved knowledge points, collect several historical learning behavior records with the same knowledge point in time sequence to obtain the historical record set of the knowledge point;

[0013] Step S103: Evaluate the mastery level of the first student for the learned knowledge points. The specific evaluation process is as follows:

[0014] Obtain the historical record set of the knowledge point, and form a time step at a time point according to each historical learning behavior record in the historical record set. Among them, the total number of time steps t of the knowledge point in the current period is the same as the total number of each historical learning behavior record T, and t = 1, 2,..., T;

[0015] Step S104: Set the eigenvalue a, and set the knowledge mastery variable of the first student for the knowledge point as L and the result variable as Q, where a > 0, L ∈ {0, a}, and Q ∈ {0, a};

[0016] Obtain the preset initial probability b, and initialize the probabilities of the mastery levels of the knowledge points as p(L0=a)=b and p(L0=0)=1-b respectively. In the prediction step, calculate the prior probability p(L t =a|Q 1:t-1 ) of the first student's mastery level of the knowledge point within time step t;

[0017] Step S105: When the answer result of the first student for the knowledge point in the historical learning behavior record corresponding to a certain time step is correct, the result variable Q=a; otherwise, Q=0. In the update step, obtain the result variable Q of the knowledge point in each time step, and calculate the posterior probability p(L t =a|Q 1:t ) of the first student's mastery level of the knowledge point within time step t, and use the posterior probability as the prior probability for the next time step;

[0018] And continuously update the prior probability of the knowledge point to obtain the characteristic mastery probability p(L T =1|Q 1:T ) of the first student for the knowledge point within the current period;

[0019] Step S106: When the characteristic mastery probability of the knowledge point is greater than the preset mastery probability threshold, it is determined that the first student has mastered the knowledge point; when the characteristic mastery probability is less than or equal to the mastery probability threshold, it is determined that the first student has not mastered the knowledge point, and the knowledge point is recorded as a defective knowledge point of the first student;

[0020] Collect the several defective knowledge points and the characteristic mastery probabilities of several knowledge points of the first student within the current period to obtain the knowledge mastery data of the first student.

[0021] Furthermore, step S200 includes:

[0022] Step S201: Obtain the knowledge mastery data of the first student, and obtain several defective knowledge points of the first student from the knowledge mastery data;

[0023] Step S202: Obtain the subject to which the defective knowledge points of the first student belong, obtain the knowledge graph of the subject, and evaluate the learning influence degree of the knowledge points in the knowledge graph on the first student's learning of the defective knowledge points. The specific process is as follows:

[0024] Obtain the upstream knowledge points directly or indirectly connected to the defective knowledge points from the knowledge graph, and record them as the precursor knowledge points of the defective knowledge points;

[0025] When the characteristic mastery probability of the precursor knowledge point is less than or equal to the mastery probability threshold, it is determined that the precursor knowledge point has a learning influence on the first student's mastery of the defective knowledge point, and the precursor knowledge point is recorded as a defective precursor knowledge point;

[0026] Step S203: obtaining several defect precursor knowledge points of the defect knowledge point, sorting them from the most upstream defect precursor knowledge point according to their positions in the knowledge graph and their connection relationships, and aggregating the several defect precursor knowledge points, taking the defect knowledge point as the last element, to obtain a defect knowledge set of the defect knowledge point;

[0027] Step S204: acquiring and aggregating defect knowledge sets of several defect knowledge points of the first student to obtain defect learning data of the first student;

[0028] In the above steps, by analyzing the students' mastery of different knowledge points in the learning process, the defective knowledge points with mastery defects are obtained. Taking into account that the mastery of knowledge is not isolated, the knowledge graph is used to find several defective knowledge points related to the defective knowledge points, and collect them to obtain the defective knowledge set of the defective knowledge points. In this way, in the process of tutoring students on defective knowledge points, only the defective knowledge set is needed to quickly and accurately obtain the students' tutoring materials, so that students can master the defective knowledge points as quickly as possible.

[0029] Furthermore, step S300 includes:

[0030] Step S301: Acquire other users who are studying on the teaching platform, and when the knowledge points learned by a certain other student in the historical period contain the knowledge points that the first student has learned, retain the certain other student;

[0031] Step S302: Obtain the retained historical learning behavior records of other students, obtain the most recent historical learning behavior record of the first student in the current period, and record it as the characteristic historical learning behavior record of the first student;

[0032] Extract several knowledge points involved in learning from the characteristic historical learning behavior records, obtain the knowledge graph in the teaching platform, obtain the knowledge point at the bottom of the knowledge graph from the several knowledge points, and record it as the characteristic knowledge point of the first student;

[0033] Step S303: Obtain the learning stage of the characteristic knowledge points that other students first learned, and record it as the characteristic learning stage, remove the historical learning behavior records of other students after the characteristic learning stage, obtain the learning time and test scores of other students from the historical learning behavior records, normalize the learning time and test scores respectively, divide the test scores by the learning time, and obtain the marked learning efficiency of other students;

[0034] Obtain the average value of the characteristic learning efficiency of several retained historical learning behavior records of other students, and denote it as the characteristic learning efficiency u' of other students. Obtain the characteristic learning efficiency u of the first student;

[0035] Step S304: Obtain the historical knowledge mastery data of other students in the characteristic learning stage. From the historical knowledge mastery data, obtain the characteristic mastery probability of the defective knowledge points of other students;

[0036] Obtain the total number n of the defective knowledge points of other students. Obtain the total number m of each defective knowledge point of the first student from the defective learning data. Calculate the defective ratio coefficient r = m / max{n, m} between the first student and other students. Obtain the defective approximation value of the first student and other students on each defective knowledge point;

[0037] Among them, the defective approximation value D of the first student and other students on the i-th defective knowledge point of the first student i = p i / max{p' i , p i}, where p i is the characteristic mastery probability of the i-th defective knowledge point of the first student in the defective learning data, and p' i is the characteristic mastery probability of the i-th defective knowledge point of other students in the historical knowledge mastery data. Obtain the average value e of the defective approximation values of the first student and other students on each defective knowledge point of the first student;

[0038] Step S305: Obtain the characteristic learning efficiency approximation value f = u / max{u, u'} of other students and the first student. Calculate the learning reference value H of other students to the first student: H = f × r × e;

[0039] When the learning reference value H is greater than the preset threshold, determine that other students have reference value for the learning of the first student, and record other students as the target reference students of the first student.

[0040] Furthermore, step S400 includes:

[0041] Step S401: Obtain each target reference student of the first student. From each target reference student, retain the target reference students whose learning rankings in the current period are in the top v% of their respective grades, where v is a preset value;

[0042] Step S402: Obtain the historical learning data of the target reference students. The historical learning data includes the learning resources of the target reference students in the historical period;

[0043] Obtain the defective knowledge set of defective knowledge points in the defective learning data of the first student. From the historical learning data of several target reference students of the first student, obtain the learning materials related to the defective knowledge points in the defective knowledge set, and use them as the teaching reference data for the defective knowledge points in the defective learning data;

[0044] Send the defective learning data and teaching reference data of the first student to the first student's class teacher on the teaching platform to assist the class teacher in teaching the first student.

[0045] In order to better implement the above method, an intelligent teaching assistance system based on behavior data and knowledge graph is also proposed. The system includes a knowledge mastery evaluation module, a defective learning data module, a reference value analysis module, and a teaching assistance module;

[0046] The knowledge mastery evaluation module is used to obtain the historical learning behavior records of students, evaluate the mastery level of the first student on different learned knowledge points, and obtain knowledge mastery data;

[0047] The defective learning data module is used to obtain the knowledge graph of the first student, evaluate the influence degree of the knowledge points in the knowledge graph on the learning of defective knowledge points by the first student, and obtain defective learning data;

[0048] The reference value analysis module is used to analyze the learning reference value of other students to the first student to obtain target reference students;

[0049] The teaching assistance module is used to generate teaching reference data for the first student, and push the teaching reference data to the first student's class teacher to assist the class teacher in teaching the first student.

[0050] Furthermore, the knowledge mastery evaluation module includes a feature mastery probability unit and a knowledge mastery evaluation unit;

[0051] The feature mastery probability unit is used to obtain the feature mastery probability of the first student on the knowledge points in the current period;

[0052] The knowledge mastery evaluation unit is used to evaluate the mastery level of the first student on different learned knowledge points according to the feature mastery probability, and obtain the knowledge mastery data of the first student.

[0053] Furthermore, the defective learning data module includes a defective knowledge set unit and a defective learning data unit;

[0054] The defective knowledge set unit is used to obtain several defective precursor knowledge points of the defective knowledge points, and collect the several defective precursor knowledge points according to the connection relationship according to their positions in the knowledge graph to obtain the defective knowledge set of the defective knowledge points;

[0055] A defect learning data unit is used to obtain and collect the defect knowledge sets of several defect knowledge points of the first student, and obtain the defect learning data of the first student.

[0056] Furthermore, the reference value analysis module includes a learning reference value unit and a reference value analysis unit;

[0057] The learning reference value unit is used to calculate the learning reference value of other students for the first student;

[0058] The reference value analysis unit is used to analyze the learning reference value of other students for the first student according to the learning reference value, and obtain the target reference students.

[0059] Furthermore, the teaching assistance module includes a teaching reference data unit and a teaching assistance unit;

[0060] The teaching reference data unit is used to obtain the defect knowledge sets of the defect knowledge points in the defect learning data of the first student, and obtain the learning materials related to the defect knowledge points in the defect knowledge sets from the historical learning data of several target reference students of the first student, and use them as the teaching reference data for the defect knowledge points in the defect learning data;

[0061] The teaching assistance unit is used to send the defect learning data and teaching reference data of the first student to the class teacher of the first student on the teaching platform to assist the class teacher in teaching the first student.

[0062] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention realizes intelligent teaching assistance for class teachers. First, according to the historical learning behaviors and knowledge graphs of students, the defect knowledge sets of defect knowledge points are obtained. In the teaching process, students only need to learn the defect knowledge sets to quickly master the defect knowledge points. And considering that other students have reference value for the current student in the historical learning process, the target reference students of the student are obtained. Only according to the historical learning materials of the target students, the teaching reference data of the student can be generated, so as to assist the teacher in teaching, improve the enthusiasm of students in learning, and effectively improve the academic performance of students. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is the method flow chart of the intelligent teaching assistance method based on behavior data and knowledge graph of the present invention;

[0064] Figure 2 is the module schematic diagram of the intelligent teaching assistance system based on behavior data and knowledge graph of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0066] Embodiment: As Figure 1 - Figure 2 shown, the present invention provides a technical solution, an intelligent teaching assistance method based on behavior data and a knowledge graph. The method includes:

[0067] Step S100: Obtain the historical learning behavior records of the first student, obtain historical learning behavior data from the historical learning behavior records, evaluate the mastery level of the first student for different learned knowledge points, and obtain knowledge mastery data;

[0068] Among them, step S100 includes:

[0069] Step S101: Monitor and record the learning behavior of the student on the teaching platform to obtain the historical learning behavior records of the first student, and extract historical learning behavior data from the historical learning behavior records. The historical learning behavior data includes several knowledge points involved in the learning behavior of the first student;

[0070] Step S102: According to the different knowledge points involved, collect several historical learning behavior records with the same knowledge point in chronological order to obtain a historical record set of the knowledge point;

[0071] Step S103: Evaluate the mastery level of the first student for the learned knowledge points. The specific evaluation process is as follows:

[0072] Obtain the historical record set of the knowledge point. According to each historical learning behavior record in the historical record set, form a time step of time points. Among them, the total number of time steps t of the knowledge point in the current period is the same as the total number of each historical learning behavior record T, and t = 1, 2,..., T;

[0073] For example, the time steps are arranged in sequence to form a sequence. The update of each time step depends on the state of the previous time step and the current answer;

[0074] Step S104: Set the eigenvalue a, set the knowledge mastery variable of the first student for the knowledge point as L, and the result variable as Q, where a > 0, L ∈ {0, a}, and Q ∈ {0, a};

[0075] For example, in the actual calculation process, generally a = 1. When L tWhen L = 1, it indicates that the first student has mastered the knowledge point at time step t. When L t = 0, it indicates that the first student has not mastered the knowledge point at time step t;

[0076] When Q t = 1, it indicates that the answer result of the first student to the question corresponding to the knowledge point at time step t is correct. When Q t = 0, it indicates that the answer result of the first student to the question corresponding to the knowledge point at time step t is wrong;

[0077] Obtain the preset initial probability b, and initialize the probability of the mastery level of the knowledge point p(L0 = a) = b, p(L0 = 0) = 1 - b. In the prediction step, calculate the prior probability p(L t = a|Q 1:t-1 ) of the first student's mastery level of the knowledge point within time step t;

[0078] For example, the prediction step predicts the current state based on the previous day. In the prediction step, the transition probability needs to be substituted to calculate the prior probability p(L t = a|Q 1:t-1 ) in time step t. Therefore, the prior probability p(L t = a|Q 1:t-1 ) = p(L t-1 = a|Q 1:t-1 ) + p(T)·p(L t-1 = 0|Q 1:t-1 ), where p(T) is the learning rate preset by the teaching platform for the first student, p(L t-1 = a|Q 1:t-1 ) is the posterior probability of the first student's mastery level of the knowledge point within time step t - 1, and p(L t-1 = 0|Q 1:t-1 ) is the posterior probability of the first student's non - mastery of the knowledge point within time step t - 1;

[0079] Step S105: When the answer result of the first student to the knowledge point in the historical learning behavior record corresponding to a certain time step is correct, the result variable Q = a; otherwise, Q = 0. In the update step, obtain the result variable Q of the knowledge point in each time step, and calculate the posterior probability p(L t = a|Q 1:t ) of the first student's mastery level of the knowledge point within time step t, and use the posterior probability as the prior probability for the next time step;

[0080] For example, in the update step, p(L t = a|Q 1:t ):

[0081] ,

[0082] Among them, p(Q t |Q 1:t-1 ) = p(Q t |L t = a)·p(L t = a|Q 1:t-1 ) + p(Q t |L t = 0)·p(L t = 0|Q 1:t-1 );p(Q t |L t = a) is the posterior probability of correctly answering the question when the first student masters the knowledge point in time step t; p(Q t |L t = 0) is the posterior probability of correctly answering the question when the first student does not master the knowledge point in time step t;

[0083] For example, when actually calculating the posterior probability recursively, it is necessary to correct according to the current answering result;

[0084] And continuously update the prior probability of the knowledge point to obtain the probability p(L T = 1|Q 1:T ) of the first student's characteristic mastery of the knowledge point in the current cycle;

[0085] Step S106: When the probability of characteristic mastery of the knowledge point is greater than the preset mastery probability threshold, it is determined that the first student has mastered the knowledge point. When the probability of characteristic mastery is less than or equal to the mastery probability threshold, it is determined that the first student has not mastered the knowledge point, and the knowledge point is recorded as the defective knowledge point of the first student;

[0086] Collect the several defective knowledge points and the probabilities of characteristic mastery of several knowledge points of the first student in the current cycle to obtain the knowledge mastery data of the first student;

[0087] For example, the knowledge graph of the first student refers to a structured representation of the first student's mastery of a certain subject knowledge. It shows in the form of a graph the knowledge points mastered by the first student, the knowledge points not mastered, and the associations between knowledge points. Taking junior high school mathematics as an example, the knowledge graph of the first student may include the following content:

[0088] Nodes (knowledge points): Numbers and algebra include integers, fractions, decimals, algebraic expressions, etc., and geometry includes points, lines, planes, angles, triangles, etc.;

[0089] Edges (connection relationships between knowledge points): "Algebraic expressions" are the upstream knowledge points of "equations", and they are connected by edges. "Algebraic expressions" are the basis for learning "equations";

[0090] Step S200: Obtain defective knowledge points from the knowledge mastery data, obtain the knowledge graph of the first student, and evaluate the degree of learning impact of the knowledge points in the knowledge graph on the first student's learning of the defective knowledge points to obtain defective learning data;

[0091] Among them, step S200 includes:

[0092] Step S201: Obtain the knowledge mastery data of the first student, and obtain several defective knowledge points of the first student from the knowledge mastery data;

[0093] Step S202: Obtain the subject to which the defective knowledge points of the first student belong, obtain the knowledge graph of the subject, and evaluate the degree of learning impact of the knowledge points in the knowledge graph on the first student's learning of the defective knowledge points. The specific process is as follows:

[0094] For example, the subjects include mathematics, English, biology, etc.;

[0095] Obtain the upstream knowledge points directly or indirectly connected to the defective knowledge points from the knowledge graph, and record them as the precursor knowledge points of the defective knowledge points;

[0096] For example, in the knowledge graph, "multiplication" is the upstream knowledge point of "factorization";

[0097] When the feature mastery probability of the precursor knowledge point is less than or equal to the mastery probability threshold, it is determined that the precursor knowledge point has a learning impact on the first student's mastery of the defective knowledge point, and the precursor knowledge point is recorded as the defective precursor knowledge point;

[0098] Step S203: Obtain several defective precursor knowledge points of the defective knowledge points, sort them according to the connection relationship starting from the most upstream defective precursor knowledge point according to their positions in the knowledge graph, and collect several defective precursor knowledge points, and use the defective knowledge point as the last element to obtain the defective knowledge set of the defective knowledge point;

[0099] Step S204: Obtain and collect the defective knowledge sets of several defective knowledge points of the first student to obtain the defective learning data of the first student;

[0100] Step S300: Obtain the historical learning behavior records and historical knowledge mastery data of other students, and analyze the learning reference value of other students to the first student in combination with the historical learning behavior records and defective learning data of the first student to obtain the target reference students;

[0101] Among them, step S300 includes:

[0102] Step S301: Obtain each other user who studies on the teaching platform. When among the various knowledge points learned by a certain other student within the historical period, there are knowledge points that the first student has already learned, retain the certain other student.

[0103] Step S302: Obtain the historical learning behavior records of the retained other students, and obtain the historical learning behavior record of the first student that is closest to the current period, and record it as the characteristic historical learning behavior record of the first student.

[0104] Extract several knowledge points involved in learning from the characteristic historical learning behavior record, obtain the knowledge graph in the teaching platform, and obtain the knowledge point that is at the most downstream in the knowledge graph from the several knowledge points, and record it as the characteristic knowledge point of the first student.

[0105] Step S303: Obtain the learning stage when the other students first learned the characteristic knowledge point, and record it as the characteristic learning stage. Eliminate the historical learning behavior records of the other students generated after the characteristic learning stage. Obtain the learning duration and test scores of the other students from the historical learning behavior records, perform normalization processing on the learning duration and test scores respectively, and divide the test score by the learning duration to obtain the marked learning efficiency of the other students.

[0106] For example, this learning stage includes the first semester of junior high school, the second semester of junior high school, etc.

[0107] Obtain the average value of the characteristic learning efficiencies of the retained several historical learning behavior records of the other students, and record it as the characteristic learning efficiency u´ of the other students. Obtain the characteristic learning efficiency u of the first student.

[0108] Step S304: Obtain the historical knowledge mastery data of the other students in the characteristic learning stage, and obtain the characteristic mastery probability of the defective knowledge points of the other students from the historical knowledge mastery data.

[0109] Obtain the total number n of the defective knowledge points of the other students, obtain the total number m of each defective knowledge point of the first student from the defective learning data, calculate the defect ratio coefficient r = m / max{n, m} between the first student and the other students, and obtain the defect approximation values of the first student and the other students on each defective knowledge point.

[0110] Among them, the defect approximation value D of the first student and the other students on the i-th defective knowledge point of the first student i = p i / max{p´ i , p i}, where p i is the characteristic mastery probability of the i-th defective knowledge point of the first student in the defective learning data, and p´ iFor other students, in the historical knowledge mastery data, it is the probability of mastering the characteristics of the i-th defective knowledge point. Obtain the average value e of the defect approximations of the first student and other students on each defective knowledge point of the first student;

[0111] Step S305: Obtain the approximate characteristic learning efficiency f = u / max{u, u´} of other students and the first student, and calculate the learning reference value H of other students for the first student: H = f × r × e;

[0112] When the learning reference value H is greater than the preset threshold, determine that other students have reference value for the learning of the first student, and record other students as the target reference students of the first student;

[0113] Step S400: Obtain the historical learning data of the target reference students, and combine it with the defective learning data of the first student to generate the teaching reference data of the first student, and push the teaching reference data to the teacher in charge of the first student to assist the teacher in teaching the first student;

[0114] Among them, step S400 includes:

[0115] Step S401: Obtain each target reference student of the first student, and from each target reference student, obtain and retain the target reference students whose learning rankings in the current period are in the top v% of their respective grades, where v is a preset value;

[0116] Step S402: Obtain the historical learning data of the target reference students. The historical learning data includes the learning resources of the target reference students in the historical period;

[0117] Obtain the defective knowledge set of the defective knowledge points in the defective learning data of the first student, and obtain the learning materials related to the defective knowledge points in the defective knowledge set from the historical learning data of several target reference students of the first student, and use them as the teaching reference data for the defective knowledge points in the defective learning data;

[0118] Send the defective learning data and teaching reference data of the first student to the teacher in charge of the first student on the teaching platform to assist the teacher in teaching the first student;

[0119] In order to better implement the above method, an intelligent teaching assistance system based on behavior data and knowledge graph is also proposed. The system includes a knowledge mastery evaluation module, a defective learning data module, a reference value analysis module, and a teaching assistance module;

[0120] The knowledge mastery evaluation module is used to obtain the historical learning behavior records of students, evaluate the mastery degree of the first student on different learned knowledge points, and obtain the knowledge mastery data;

[0121] A defect learning data module, which is used to obtain the knowledge graph of the first student, evaluate the learning influence degree of the knowledge points in the knowledge graph on the learning of the defective knowledge points of the first student, and obtain defect learning data;

[0122] A reference value analysis module, which is used to analyze the learning reference value of other students to the first student and obtain the target reference students;

[0123] A teaching assistance module, which is used to generate teaching reference data for the first student, push the teaching reference data to the teacher of the first student, and assist the teacher in teaching the first student;

[0124] Among them, the knowledge mastery evaluation module includes a feature mastery probability unit and a knowledge mastery evaluation unit;

[0125] The feature mastery probability unit is used to obtain the feature mastery probability of the first student for the knowledge points in the current period;

[0126] The knowledge mastery evaluation unit is used to evaluate the mastery degree of the first student for different learned knowledge points according to the feature mastery probability, and obtain the knowledge mastery data of the first student.

[0127] Among them, the defect learning data module includes a defective knowledge set unit and a defect learning data unit;

[0128] The defective knowledge set unit is used to obtain several defective precursor knowledge points of the defective knowledge points, and collect the several defective precursor knowledge points according to the connection relationship according to their positions in the knowledge graph, and obtain the defective knowledge set of the defective knowledge points;

[0129] The defect learning data unit is used to obtain and collect the defective knowledge sets of several defective knowledge points of the first student, and obtain the defect learning data of the first student.

[0130] Among them, the reference value analysis module includes a learning reference value unit and a reference value analysis unit;

[0131] The learning reference value unit is used to calculate the learning reference value of other students to the first student;

[0132] The reference value analysis unit is used to analyze the learning reference value of other students to the first student according to the learning reference value, and obtain the target reference students;

[0133] Among them, the teaching assistance module includes a teaching reference data unit and a teaching assistance unit;

[0134] A teaching reference data unit is used to obtain a set of defective knowledge points of defective learning data of the first student, and obtain learning materials related to the defective knowledge points within the set of defective knowledge from the historical learning data of a number of target reference students of the first student, and use them as teaching reference data for the defective knowledge points in the defective learning data;

[0135] A teaching assistance unit is used to send the defective learning data and teaching reference data of the first student to the teacher in charge of the first student on the teaching platform to assist the teacher in charge in teaching the first student.

[0136] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. An intelligent teaching assistance method based on behavioral data and knowledge graph, characterized in that: The method comprises: Step S100: obtaining a historical learning behavior record of a first student, obtaining historical learning behavior data from the historical learning behavior record, evaluating the first student's mastery of different knowledge points learned, and obtaining knowledge mastery data; Step S200: obtaining defective knowledge points from the knowledge mastery data, obtaining the knowledge graph of the first student, evaluating the influence of the knowledge points in the knowledge graph on the first student's learning of the defective knowledge points, and obtaining defective learning data; Step S300: Obtaining the historical learning behavior records and historical knowledge mastery data of other students, and combining the historical learning behavior records and defective learning data of the first student, analyzing the learning reference value of the other students to the first student, and obtaining the target reference student; Step S400: acquiring the historical learning data of the target reference student, and combining it with the defective learning data of the first student to generate teaching reference data of the first student, and pushing the teaching reference data to the teacher of the first student to assist the teacher in teaching the first student; The step S400 includes: Step S401: Obtain target reference students of the first student, and from among the target reference students, obtain and retain target reference students whose learning rankings are in the top v% of their grade in the current cycle, where v is a preset value; Step S402: Acquire the historical learning data of the target reference student, wherein the historical learning data includes the learning resources of the target reference student in a historical period; Obtaining a defective knowledge set of defective knowledge points in the defective learning data of the first student, and obtaining learning materials involving defective knowledge points in the defective knowledge set from historical learning data of several target reference students of the first student, and using the learning materials as teaching reference data for the defective knowledge points in the defective learning data; The defective learning data and teaching reference data of the first student are sent to the teacher of the first student on the teaching platform to assist the teacher in teaching the first student.

2. The intelligent teaching assistance method based on behavior data and knowledge graph according to claim 1 is characterized in that: The step S100 includes: Step S101: monitoring and recording the learning behavior of the student on the teaching platform to obtain a historical learning behavior record of the first student, and extracting historical learning behavior data from the historical learning behavior record, wherein the historical learning behavior data includes a number of knowledge points involved in the learning behavior of the first student; Step S102: according to different knowledge points involved, a plurality of historical learning behavior records of the same knowledge point are collected in chronological order to obtain a historical record set of the knowledge point; Step S103: Evaluate the first student's mastery of the learned knowledge points. The specific evaluation process is as follows: Obtain a historical record set of the knowledge point, and form a time step of the time point according to each historical learning behavior record in the historical record set, wherein the total number of time steps t of the knowledge point in the current cycle is the same as the total number T of each historical learning behavior record, t=1, 2, ..., T; Step S104: setting a characteristic value a, setting the first student's knowledge mastery variable of the knowledge point to L, and the result variable to Q, where a>0, L∈{0,a}, Q∈{0,a}; Obtain the preset initial probability b, initialize the probabilities of the mastery of the knowledge point p(L0=a)=b and p(L0=0)=1-b respectively, and in the prediction step, calculate the prior probability p(L0) of the mastery of the knowledge point by the first student within time step t. t =a|Q 1:t-1 ); Step S105: When the answer result of the first student to the knowledge point in the historical learning behavior record corresponding to a certain time step is correct, the result variable Q=a, otherwise Q=0. In the update step, the result variable Q of the knowledge point in each time step is obtained, and the posterior probability p(L) of the first student's mastery of the knowledge point in the time step t is calculated. t =a|Q 1:t ), taking the posterior probability as the prior probability of the next time step; The prior probability of the knowledge point is continuously updated to obtain the probability p(L T =1|Q 1:T ); Step S106: when the characteristic mastering probability of the knowledge point is greater than a preset mastering probability threshold, it is determined that the first student has mastered the knowledge point; when the characteristic mastering probability is less than or equal to the mastering probability threshold, it is determined that the first student has not mastered the knowledge point, and the knowledge point is recorded as a defective knowledge point of the first student; The plurality of defective knowledge points and the characteristic mastering probabilities of a plurality of knowledge points of the first student in the current cycle are collected to obtain the knowledge mastering data of the first student.

3. The intelligent teaching assistance method based on behavior data and knowledge graph according to claim 2 is characterized in that: The step S200 includes: Step S201: obtaining the knowledge mastery data of the first student, and obtaining a number of deficient knowledge points of the first student from the knowledge mastery data; Step S202: Obtain the subject to which the deficient knowledge point of the first student belongs, obtain the knowledge graph of the subject, and evaluate the influence of the knowledge points in the knowledge graph on the first student's learning of the deficient knowledge point. The specific process is as follows: Acquire upstream knowledge points directly or indirectly connected to the defect knowledge point from the knowledge graph, and record them as predecessor knowledge points of the defect knowledge point; When the characteristic mastering probability of the precursor knowledge point is less than or equal to the mastering probability threshold, it is determined that the precursor knowledge point has a learning influence on the first student's mastering of the defective knowledge point, and the precursor knowledge point is recorded as a defective precursor knowledge point; Step S203: obtaining several defect precursor knowledge points of the defect knowledge point, sorting the several defect precursor knowledge points from the upstream defect precursor knowledge point according to their positions in the knowledge graph and their connection relationships, and aggregating the several defect precursor knowledge points, taking the defect knowledge point as the last element, to obtain a defect knowledge set of the defect knowledge point; Step S204: acquiring and aggregating defect knowledge sets of several defect knowledge points of the first student to obtain defect learning data of the first student.

4. The intelligent teaching assistance method based on behavior data and knowledge graph according to claim 3 is characterized in that: The step S300 includes: Step S301: Acquire other users who are studying on the teaching platform, and when the knowledge points studied by a certain other student in the historical period contain the knowledge points that the first student has studied, retain the certain other student; Step S302: Obtain the retained historical learning behavior records of other students, obtain the historical learning behavior record of the first student closest to the current period, and record it as the characteristic historical learning behavior record of the first student; Extracting several knowledge points involved in learning from the characteristic historical learning behavior records, obtaining a knowledge graph in the teaching platform, and obtaining the knowledge point at the bottom of the knowledge graph from the several knowledge points, and recording it as the characteristic knowledge point of the first student; Step S303: obtaining the learning stage when other students first learn the characteristic knowledge point, and recording it as the characteristic learning stage, removing the historical learning behavior records of the other students generated after the characteristic learning stage, obtaining the learning time and test scores of the other students from the historical learning behavior records, normalizing the learning time and test scores respectively, dividing the test scores by the learning time, and obtaining the marked learning efficiency of the other students; Obtain an average value of the characteristic learning efficiency of several retained historical learning behavior records of the other students, and record it as the characteristic learning efficiency u´ of the other students, and obtain the characteristic learning efficiency u of the first student; Step S304: obtaining the historical knowledge mastery data of the other students in the feature learning stage, and obtaining the feature mastery probability of the defective knowledge points of the other students from the historical knowledge mastery data; Obtain the total number of defective knowledge points n of the other students, obtain the total number of defective knowledge points m of the first student from the defective learning data, calculate the defect ratio coefficient r=m / max{n,m} between the first student and the other students, and obtain the defect approximate value of the first student and the other students on the defective knowledge points; Among them, the defect approximate value D of the first student and the other students on the i-th defective knowledge point of the first student i =p i / max{p´ i ,p i }, where p i is the probability of mastering the features of the i-th defective knowledge point in the defective learning data of the first student, p´ i For the feature mastery probability of the i-th defective knowledge point in the historical knowledge mastery data of the other students, obtain the average value e of the defect approximation values ​​of the first student and the other students on each defective knowledge point of the first student; Step S305: obtaining an approximate characteristic learning efficiency f=u / max{u,u´} of the other students and the first student, and calculating a learning reference value H=f×r×e of the other students to the first student; When the learning reference value H is greater than a preset threshold, it is determined that the other students have reference value to the first student's learning, and the other students are recorded as target reference students for the first student.

5. An intelligent teaching assistance system based on behavior data and knowledge graph, used to execute the intelligent teaching assistance method based on behavior data and knowledge graph as described in any one of claims 1 to 4, characterized in that: The system includes a knowledge mastery assessment module, a defect learning data module, a reference value analysis module, and a teaching assistance module; The knowledge mastery evaluation module is used to obtain the student's historical learning behavior records, evaluate the first student's mastery of different knowledge points learned, and obtain knowledge mastery data; The defective learning data module is used to obtain the knowledge graph of the first student, evaluate the influence of the knowledge points in the knowledge graph on the learning of the defective knowledge points by the first student, and obtain defective learning data; The reference value analysis module is used to analyze the learning reference value of the other students to the first student to obtain the target reference student; The teaching auxiliary module is used to generate teaching reference data for the first student, push the teaching reference data to the teacher of the first student, and assist the teacher in teaching the first student.

6. The intelligent teaching assistance system based on behavior data and knowledge graph according to claim 5 is characterized in that: The knowledge mastery assessment module includes a feature mastery probability unit and a knowledge mastery assessment unit; The feature mastering probability unit is used to obtain the feature mastering probability of the first student for the knowledge point in the current cycle; The knowledge mastery evaluation unit is used to evaluate the mastery degree of the first student on different knowledge points learned according to the feature mastery probability, and obtain the knowledge mastery data of the first student.

7. The intelligent teaching assistance system based on behavior data and knowledge graph according to claim 5 is characterized in that: The defect learning data module includes a defect knowledge set unit and a defect learning data unit; The defect knowledge set unit is used to obtain several defect precursor knowledge points of the defect knowledge point, and according to the positions of the several defect precursor knowledge points in the knowledge graph and the connection relationship, the several defect precursor knowledge points are collected to obtain the defect knowledge set of the defect knowledge point; The defect learning data unit is used to acquire and aggregate the defect knowledge sets of several defect knowledge points of the first student to obtain the defect learning data of the first student.

8. The intelligent teaching assistance system based on behavior data and knowledge graph according to claim 5 is characterized in that: The reference value analysis module includes a learning reference value unit and a reference value analysis unit; The learning reference value unit is used to calculate the learning reference value of the other students to the first student; The reference value analysis unit is used to analyze the learning reference values ​​of the other students to the first student according to the learning reference value to obtain a target reference student.

9. The intelligent teaching assistance system based on behavior data and knowledge graph according to claim 5, characterized in that: The teaching auxiliary module includes a teaching reference data unit and a teaching auxiliary unit; The teaching reference data unit is used to obtain a defective knowledge set of defective knowledge points in the defective learning data of the first student, and to obtain learning materials involving defective knowledge points in the defective knowledge set from historical learning data of several target reference students of the first student, and to use the learning materials as teaching reference data for the defective knowledge points in the defective learning data; The teaching auxiliary unit is used to send the defect learning data and teaching reference data of the first student to the teacher of the first student on the teaching platform, so as to assist the teacher in teaching the first student.

Images