Method for evaluating knowledge mastering level of student and pushing corresponding learning content
By constructing a knowledge question bank and using Rasch model to evaluate students' mastery level, and generating personalized learning content push strategies, the problem that existing platforms cannot accurately evaluate students' mastery of knowledge points is solved, and quantitative analysis and the formulation of personalized learning plans are realized.
Patent Information
- Application Number
- CN202510519182.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-24
- Publication Date
- 2025-08-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing intelligent learning platform is difficult to accurately evaluate students' mastery of various knowledge points, resulting in the inability to formulate personalized learning plans for different individual students.
A question bank containing N knowledge questions is constructed, a bipartite data of students is fitted through the Rasch model, a threshold for correct answers is set, a personalized learning content push strategy is generated, and the number of test questions and learning content is adjusted according to the students' mastery level.
It realizes quantitative analysis of students' knowledge mastery, accurately locates students' specific mastery of each knowledge point, and can formulate personalized learning plans for different individual students, and push targeted learning content.
Smart Images
Figure CN120470111A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of smart education technology, and in particular to a method for evaluating students' knowledge mastery levels and pushing corresponding learning content. Background Art
[0002] With the continuous development of educational technology, intelligent learning platforms have gradually become an important tool in the education field. These platforms integrate technologies such as artificial intelligence, big data analysis, and natural language processing to provide students with personalized learning experiences. For example, some platforms use dynamic knowledge tracking technologies (such as Bayesian knowledge tracking (BKT)) to update students' mastery of knowledge points in real time and recommend appropriate learning content based on this data. In addition, through natural language processing and deep learning technologies, the platforms can understand students' natural language input and generate personalized learning paths and content.
[0003] However, most current intelligent learning platforms still have limitations when it comes to assessing students' knowledge mastery. Traditional assessment methods rely primarily on test scores, a subjective scoring method that is unable to quantify students' knowledge mastery. Although some platforms have begun to experiment with using big data analysis and machine learning algorithms to assess students' learning, these methods often only provide relatively general analyses, making it difficult to accurately pinpoint students' specific mastery of various knowledge points, and thus unable to develop specific learning plans for different students. To this end, this paper proposes a method based on assessing students' knowledge mastery and delivering corresponding learning content. Summary of the Invention
[0004] The main purpose of the present invention is to provide a method based on evaluating students' knowledge mastery level and pushing corresponding learning content, which can effectively solve the problems in the background technology.
[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:
[0006] Methods based on assessing students' knowledge mastery and delivering corresponding learning content include:
[0007] Step 1: Construct a knowledge point question bank, wherein the test question bank contains N knowledge point question sets, and any j-th knowledge point question set contains at least one test question for evaluating the mastery of the j-th knowledge point, where j = 1, 2, ..., N; randomly combine the test questions with a period of λ, and generate a test question set for the evaluated student i based on the combination results, where i = 1, 2, ..., M, where M is the number of students to be evaluated;
[0008] Step 2: Construct a knowledge point learning content library, wherein the knowledge point learning content library contains N knowledge point training sets, and any j-th knowledge point training set contains at least one learning content push information related to the j-th knowledge point;
[0009] Step 3: collecting binary data of the student i's responses to all test questions in the test set, wherein the binary data is determined based on the student i's responses to the test questions, with a correct answer being recorded as 1 and an incorrect answer being recorded as 0;
[0010] Step 4: Use the Rasch model to fit the obtained binary data to obtain the correct answer probability P, which reflects the degree of mastery of the jth knowledge point by the assessed student i. ij , set the correct answer probability classification threshold of the jth knowledge point to P j1 and P j2 , using the classification threshold, the mastery level of the assessed student i on the jth knowledge point is classified into fully mastered, basically mastered and not yet mastered;
[0011] Step 5: Generate a set of learning content push information for the assessed student i based on the assessed student i's mastery of the jth knowledge point, and form a push strategy i.
[0012] The method further comprises:
[0013] According to the current mastery of the jth knowledge point by the assessed student i, a strategy for generating the test question set for the assessed student i in the next cycle is formulated.
[0014] Furthermore, the probability of correct answer P ij The calculation formula is:
[0015]
[0016] Where θ i is the answering ability coefficient of the assessed student i; β j is the difficulty coefficient of the j-th knowledge point; e is the base of the natural logarithm.
[0017] Furthermore, the generation strategy of the test question set for the evaluated student i in the next cycle is as follows:
[0018] When the assessed student i has fully mastered the j-th knowledge point, the number of test questions used to assess the mastery of the j-th knowledge point in the test question set for the assessed student i in the next cycle is reduced;
[0019] When the mastery level of the j-th knowledge point of the assessed student i is not yet mastered, the number of test questions used to assess the mastery level of the j-th knowledge point in the test question set for the assessed student i in the next cycle is increased;
[0020] When the assessed student i has basically mastered the j-th knowledge point, the number of test questions used to assess the mastery of the j-th knowledge point in the test question set for the assessed student i in the next cycle remains unchanged.
[0021] Furthermore, in step 5, the specific content of the push strategy i includes:
[0022] When the assessed student i has fully mastered the jth knowledge point, reduce the learning content push information for the assessed student i and focus on the learning content push information related to the jth knowledge point;
[0023] When the assessed student i's mastery level of the j-th knowledge point is not yet mastered, the learning content push information related to the j-th knowledge point is added to the learning content push information set for the assessed student i;
[0024] When the assessed student i has basically mastered the j-th knowledge point, the number of learning content push information related to the j-th knowledge point in the learning content push information set for the assessed student i is maintained unchanged.
[0025] Furthermore, the learning content push information set includes at least one piece of learning content push information related to the j-th knowledge point.
[0026] Furthermore, the test question set includes at least one test question for evaluating the mastery of the j-th knowledge point.
[0027] Furthermore, the classification principle for the mastery level of the jth knowledge point of the evaluated student i is:
[0028] When P ij <P j1 When , the level of mastery of the jth knowledge point by the assessed student i is not yet mastered;
[0029] When P ij >P j2 When , the assessed student i’s mastery level of the jth knowledge point is complete;
[0030] When P j1 ≤P ij ≤P j2 When , the assessed student i’s mastery level of the j-th knowledge point is basic mastery.
[0031] Furthermore, P j1 and Pj2 The value ranges are: P j1 ∈[0.75,0.9]; P j2 ∈[0.95,1).
[0032] The present invention has the following beneficial effects:
[0033] Compared with the prior art, a knowledge point question bank comprising N knowledge point question sets is constructed, wherein any j-th knowledge point question set comprises at least one test question for evaluating the mastery of the j-th knowledge point, the test questions are randomly combined with a period of λ, and a test question set for the assessed student i is generated according to the combination result. A knowledge point learning content bank comprising N knowledge point training sets is constructed, wherein any j-th knowledge point training set comprises at least one learning content push information related to the j-th knowledge point, binary data after the assessed student i answers all the test questions in the test question set are collected, and the obtained binary data are fitted using a Rasch model to obtain a correct answer probability P for reflecting the mastery of the assessed student i on the j-th knowledge point. ij , set the correct answer probability classification threshold of the jth knowledge point to P j1 and P j2 , use the classification threshold to classify the mastery level of the jth knowledge point of the evaluated student i, generate a set of learning content push information for the evaluated student i, and form a push strategy. By converting the student test results into quantitative data, a quantitative analysis of the student's knowledge mastery level is achieved, thereby accurately locating the specific mastery of different students on each knowledge point, so as to formulate corresponding learning plans for different students and push targeted learning content information. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 This is a flow chart of a method for evaluating students' knowledge mastery level and pushing corresponding learning content according to the present invention;
[0035] Figure 2 Schematic diagram of the composition structure of the knowledge point question bank in the technical solution of the present invention;
[0036] Figure 3 This is a schematic diagram of the composition structure of the knowledge point learning content library in the technical solution of the present invention. DETAILED DESCRIPTION
[0037] The present invention will be further described below in conjunction with specific embodiments. The accompanying drawings are for illustrative purposes only and represent only schematic diagrams rather than actual drawings. They should not be understood as limiting the present invention. In order to better illustrate the specific embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product.
[0038] In this embodiment, the process of students learning on the intelligent learning network platform through mobile terminals is taken as an example to illustrate this solution. The specific process is as follows:
[0039] Step a: Students obtain learning content (i.e., obtain push information about learning content related to knowledge points) on the intelligent learning network platform through mobile terminals;
[0040] Step b: The intelligent learning network platform evaluates students’ knowledge mastery level;
[0041] Step c: The intelligent learning network platform pushes customized learning content to students based on the assessment results of their knowledge mastery level.
[0042] The specific implementation process of the technical solution of the present invention includes the following steps:
[0043] Step 1: Build a knowledge point question bank
[0044] It should be noted that the test question bank contains N sets of knowledge point questions, and any j-th set of knowledge point questions contains at least one test question for evaluating the mastery of the j-th knowledge point, j = 1, 2, ..., N; Specifically, the structure of the knowledge point question bank is as follows: Figure 2 shown.
[0045] Step 2: Generate a test set
[0046] The test questions are randomly combined with a period of λ, and a test question set for the evaluated student i is generated based on the combination results, where i = 1, 2, ..., M; M is the number of evaluated students;
[0047] It should be noted that the generated test question set contains at least one test question for evaluating the mastery of the j-th knowledge point. For the first generated test question set, the number of test questions for any j-th knowledge point can be set randomly, but should be an integer greater than 1.
[0048] Step 3: Build a knowledge point learning content library
[0049] It should be noted that the knowledge point learning content library contains N knowledge point training sets, and any j-th knowledge point training set contains at least one learning content push information related to the j-th knowledge point; specifically, the structure of the knowledge point learning content library is as follows: Figure 3 shown.
[0050] Step 4: Collect binary data after the student i answers all the test questions in the test question set. The binary data is determined based on the student i's answer to the test question. If the answer is correct, it is recorded as 1, and if the answer is incorrect, it is recorded as 0.
[0051] The collected data can be recorded in a table or matrix format. When recorded in a matrix format, the rows of the matrix represent the students being evaluated, the columns of the matrix represent all the test questions in the test question set, and the elements in the matrix represent the responses of the students being evaluated to the test questions. Specifically, when the answer is correct, the element is 1, and when the answer is incorrect, the element is 0.
[0052] Through the above steps, the evaluation process of the students' mastery of knowledge points can be transformed into a quantitative data evaluation process.
[0053] Step 5: Use the Rasch model to fit the obtained binary data to obtain the correct answer probability P, which reflects the degree of mastery of the jth knowledge point by the assessed student i. ij , the calculation formula is:
[0054]
[0055] Where θ i is the answering ability coefficient of the assessed student i; β j is the difficulty coefficient of the j-th knowledge point; e is the base of the natural logarithm.
[0056] It should be noted that for the answering ability coefficient θ i and the difficulty coefficient β of the knowledge point j The maximum likelihood method can be used to estimate , and the specific steps are as follows:
[0057] Step S51: Construct likelihood function L
[0058] Among them, the expression of the likelihood function L can be:
[0059]
[0060] Where, δ ij It is represented as the answer result of the assessed student i to the jth knowledge point, where δ ij =0 or 1;
[0061] Step S52: Parameter estimation
[0062] The capability coefficient θ is gradually approximated by an iterative algorithm i and the difficulty coefficient β of the knowledge point j , so that the likelihood function L is maximized. Initially, the answering ability coefficient θ can be i and the difficulty coefficient β of the knowledge point jAssign arbitrary values and then iteratively adjust the coefficients until the value that maximizes the likelihood function L is found. In practice, specialized Rasch model analysis software, such as Winsteps or RUMM2030, is often used. These software programs can automate the parameter estimation process. For example, in Winsteps, users can import a data file, select an appropriate model (such as the Rasch partial credit model), and run the analysis.
[0063] Step S53: Result verification
[0064] After the coefficients are estimated, it is necessary to check the fit of the model. This can be done by looking at fit statistics such as the chi-square test, AIC, BIC, etc. If the fit is poor, the model or data may need to be refitted.
[0065] Through the above steps, the ability coefficient θ in the Rasch model can be accurately determined i and the difficulty coefficient β of the knowledge point j , thereby obtaining the correct answer probability P that reflects the mastery of the jth knowledge point by the assessed student i ij .
[0066] Step 6: Set the correct answer probability classification threshold of the jth knowledge point to P j1 and P j2 , where P j1 and P j2 The value ranges are: P j1 ∈[0.75,0.9]; P j2 ∈[0.95,1).
[0067] Step 7: Use the classification threshold to classify the mastery level of the assessed student i on the jth knowledge point into fully mastered, basically mastered, and not yet mastered. The classification principle is:
[0068] When P ij <P j1 When , the level of mastery of the jth knowledge point by the assessed student i is not yet mastered;
[0069] When P ij >P j2 When , the assessed student i’s mastery level of the jth knowledge point is complete;
[0070] When P j1 ≤P ij ≤P j2 When , the assessed student i’s mastery level of the j-th knowledge point is basic mastery.
[0071] Step 8: Generate a set of learning content push information for the assessed student i based on the assessed student i's mastery of the jth knowledge point, and form a push strategy i.
[0072] Specifically, the specific content of push strategy i includes:
[0073] Push strategy 1: When the assessed student i has fully mastered the jth knowledge point, reduce the learning content push information for the assessed student i and focus on the learning content push information related to the jth knowledge point;
[0074] Push strategy 2: When the assessed student i's mastery of the jth knowledge point is not yet mastered, the learning content push information related to the jth knowledge point is added to the learning content push information set for the assessed student i;
[0075] Push strategy 3: When the assessed student i has a basic grasp of the j-th knowledge point, the number of learning content push information related to the j-th knowledge point in the learning content push information set for the assessed student i remains unchanged.
[0076] It should be noted that in push strategy 1, after reducing the learning content push information related to the jth knowledge point in the learning content push information set for assessed student i, the learning content push information set should still include at least one learning content push information related to the jth knowledge point. This is because, although assessed student i has fully mastered the jth knowledge point, at least one learning content push information related to the jth knowledge point is still retained to prevent assessed student i from forgetting the jth knowledge point after a push cycle λ. This strengthens the memory of the jth knowledge point, allowing for a more secure grasp of the knowledge point.
[0077] Step 9: Based on the current mastery of the jth knowledge point by the assessed student i, a strategy for generating the test set for the assessed student i in the next cycle is formulated. Specifically:
[0078] Generation strategy 1: When the assessed student i has fully mastered the j-th knowledge point, reduce the number of test questions used to assess the mastery of the j-th knowledge point in the test question set for the assessed student i in the next cycle;
[0079] Generation strategy 2: When the mastery level of the j-th knowledge point of the evaluated student i is not yet mastered, increase the number of test questions used to evaluate the mastery level of the j-th knowledge point in the test question set for the evaluated student i in the next cycle;
[0080] Generation strategy 3: When the assessed student i has a basic grasp of the j-th knowledge point, the number of test questions used to assess the mastery of the j-th knowledge point in the test question set for the assessed student i in the next cycle remains unchanged.
[0081] It should be noted that in generation strategy 1, after reducing the number of test questions used to assess the mastery of the j-th knowledge point in the test set for student i in the next cycle, the generated test set should still include at least one test question used to assess the mastery of the j-th knowledge point. The reasoning is similar to that of push strategy 1 in step 8: although student i has fully mastered the j-th knowledge point, at least one test question related to the j-th knowledge point is retained to prevent student i from forgetting the j-th knowledge point after a push cycle λ. This strengthens the memory of the j-th knowledge point, allowing for more secure mastery of the knowledge point.
[0082] Through the above steps, the student test results can be converted into quantitative data, and a quantitative analysis of the students' knowledge mastery can be achieved, thereby accurately locating the specific mastery of each knowledge point by different individual students, so as to formulate corresponding learning plans for different individual students and push targeted learning content information.
[0083] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.
Claims
1. A method based on assessing students' knowledge mastery level and pushing corresponding learning content, characterized by: include: Step 1: Construct a knowledge point question bank, wherein the test question bank contains N knowledge point question sets, and any j-th knowledge point question set contains at least one test question for evaluating the mastery of the j-th knowledge point, where j = 1, 2, ..., N; randomly combine the test questions with a period of λ, and generate a test question set for the evaluated student i based on the combination results, where i = 1, 2, ..., M, where M is the number of students to be evaluated; Step 2: Construct a knowledge point learning content library, wherein the knowledge point learning content library contains N knowledge point training sets, and any j-th knowledge point training set contains at least one learning content push information related to the j-th knowledge point; Step 3: collecting binary data of the student i's responses to all test questions in the test set, wherein the binary data is determined based on the student i's responses to the test questions, with a correct answer being recorded as 1 and an incorrect answer being recorded as 0; Step 4: Use the Rasch model to fit the obtained binary data to obtain the correct answer probability P, which reflects the degree of mastery of the jth knowledge point by the assessed student i. ij , set the correct answer probability classification threshold of the jth knowledge point to P j1 and P j2 , using the classification threshold, the mastery level of the assessed student i on the jth knowledge point is classified into fully mastered, basically mastered and not yet mastered; Step 5: Generate a set of learning content push information for the assessed student i based on the assessed student i's mastery of the jth knowledge point, and form a push strategy i.
2. The method according to claim 1, characterized in that: Probability of correct answer P ij The calculation formula is: Where θ i is the answering ability coefficient of the assessed student i; β j is the difficulty coefficient of the j-th knowledge point; e is the base of the natural logarithm.
3. The method according to claim 1, characterized in that: The method further comprises: Based on the current mastery of the jth knowledge point by the assessed student i, a strategy for generating the test set for the assessed student i in the next cycle is formulated as follows: When the assessed student i has fully mastered the j-th knowledge point, the number of test questions used to assess the mastery of the j-th knowledge point in the test question set for the assessed student i in the next cycle is reduced; When the mastery level of the j-th knowledge point of the assessed student i is not yet mastered, the number of test questions used to assess the mastery level of the j-th knowledge point in the test question set for the assessed student i in the next cycle is increased; When the assessed student i has basically mastered the j-th knowledge point, the number of test questions used to assess the mastery of the j-th knowledge point in the test question set for the assessed student i in the next cycle remains unchanged.
4. The method according to claim 1, characterized in that: In step 5, the specific content of push strategy i includes: When the assessed student i has fully mastered the jth knowledge point, reduce the learning content push information for the assessed student i and focus on the learning content push information related to the jth knowledge point; When the assessed student i's mastery level of the j-th knowledge point is not yet mastered, the learning content push information related to the j-th knowledge point is added to the learning content push information set for the assessed student i; When the assessed student i has basically mastered the j-th knowledge point, the number of learning content push information related to the j-th knowledge point in the learning content push information set for the assessed student i is maintained unchanged.
5. The method according to claim 4, characterized in that: The learning content push information set includes at least one piece of learning content push information related to the j-th knowledge point.
6. The method of assessing students' knowledge mastery and pushing corresponding learning content according to claim 3, characterized in that: The test question set includes at least one test question for evaluating the mastery of the j-th knowledge point.
7. The method according to claim 1, characterized in that: The classification principle for the mastery level of the jth knowledge point of the assessed student i is: When P ij <P j1 When , the level of mastery of the jth knowledge point by the assessed student i is not yet mastered; When P ij >P j2 When , the assessed student i’s mastery level of the jth knowledge point is complete; When P j1 ≤P ij ≤P j2 When , the assessed student i’s mastery level of the j-th knowledge point is basic mastery.
8. The method of assessing students' knowledge mastery and pushing corresponding learning content according to claim 7, characterized in that: P j1 and P j2 The value ranges are: P j1 ∈[0.75,0.9]； P j2 ∈[0.95,1)。