Student cognitive modeling method based on auto-encoder-like non-negative matrix co-factorization
Through the student cognitive modeling method based on non-negative matrix cofactor decomposition, combined with low-dimensional hidden space representation and block coordinate descent projection gradient method, the problem of student achievement prediction and knowledge level diagnosis is solved, and the mastery of students' knowledge points is achieved In-depth understanding of the situation and prediction of future answer performance improve diagnostic accuracy and consistency.
Patent Information
- Application Number
- CN202410992370.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-07-23
AI Technical Summary
The existing student cognitive modeling methods have a separation between student achievement prediction and knowledge level diagnosis, and it is difficult to achieve high-precision prediction and in-depth understanding of students' knowledge level at the same time, resulting in weak diagnostic abilities and rough granularity.
Using the student cognitive modeling method based on non-negative matrix cofactor decomposition, the encoding and decoding module is constructed by obtaining students' historical answer data and exercise-knowledge point correlation information, and the low-dimensional hidden space represents the degree of students' mastery of knowledge points. , and model training is carried out in combination with the block coordinate descent projection gradient method with constraints and the Liphiz step size solution strategy to achieve the unity of student achievement prediction and knowledge level diagnosis.
It has achieved the comprehensive diagnosis of students' knowledge point mastery and future answer performance while ensuring the accuracy of students' performance prediction, breaking through the bottleneck that traditional educational data mining technology cannot deeply understand students' knowledge level, and improving the accuracy and consistency of knowledge level diagnosis.
Smart Images

Figure CN119026676A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of educational data mining, and in particular to a student cognitive modeling method based on a non-negative matrix co-factor decomposition framework. Background Art
[0002] Student cognitive modeling aims to evaluate students’ cognitive status at different stages of their learning process, specifically involving tasks such as predicting students’ academic performance and diagnosing their knowledge level. Figure 2 Establishing an accurate and reliable student cognitive model is not only an important means to achieve personalized learning, but also provides important support for building a high-quality education system.
[0003] Around the basic task of student cognitive modeling, relevant analysis methods can be divided into two classic routes. The first route is the educational psychology method represented by cognitive diagnosis models (CDMs), including continuous CDMs typified by item response theory and discrete CDMs typified by deterministic inputs, noisy "and" gate. Although cognitive diagnostic models can perform knowledge level diagnosis tasks, they are heavily dependent on predefined cognitive variables driven by expert experience, which makes it difficult to fully capture students' cognitive styles, and then perform poorly in student performance prediction tasks.
[0004] The second route is the educational data mining technology with matrix decomposition method as the core. Compared with the cognitive diagnosis model, the matrix decomposition method can directly decompose the students' answer data into a series of cognitive variables and perform the student performance prediction task. Furthermore, in order to improve the interpretability of the decomposed cognitive variables for the knowledge level diagnosis task, a scalable nonnegative matrix co-factorization model (SNMCF) uses a covering function to estimate the students' knowledge level. However, the SMNCF model separates the student performance prediction task and the knowledge level diagnosis task, that is, in the modeling process, it only focuses on the prediction effect of improving students' performance, and does not show the expression of students' mastery of knowledge points. Then, it lacks the ability to simulate the consistency of students' knowledge level and the quality of answer results, resulting in weak diagnostic ability and coarse diagnostic granularity.
[0005] Based on this, educational psychology methods can provide more accurate results for diagnosing knowledge levels, but the accuracy of predicting student performance is not enough; conversely, educational data mining technology can provide more accurate results for predicting student performance, but the diagnostic effect on students' knowledge levels is not good. Summary of the invention
[0006] The purpose of the present invention is to propose a student cognitive modeling method based on a non-negative matrix co-factorization framework. From the perspective of educational data mining technology, while ensuring high-precision prediction of student grades, it breaks through the bottleneck that educational data mining technology cannot deeply understand students' knowledge level. It can comprehensively diagnose students' current knowledge mastery and predict their future answer performance.
[0007] To achieve the above object, the technical solution of the present invention is: a student cognitive modeling method based on autoencoder-like non-negative matrix co-factor decomposition, comprising the following steps:
[0008] Step 1: Obtain students’ historical answer data and extract the question-knowledge point association information in the test text;
[0009] Step 2: Based on the students’ answer data and the relationship between exercises and knowledge points, a coding module is built to construct a low-dimensional latent space representation of students, exercises, and knowledge points;
[0010] Step 3: Use the low-dimensional latent space information of students, exercises and knowledge points to build a decoding module to indicate the students' mastery of knowledge points;
[0011] Step 4: Integrate the encoding module and the decoding module to complete the construction of the student cognitive model;
[0012] Step 5: Use the constrained block coordinate descent projected gradient method and combine it with the Lipschitz step size solution strategy to train the model;
[0013] Step 6: Complete the learning performance prediction task based on the exercises and the low-dimensional latent space representation of the students, and complete the knowledge level diagnosis task based on the students' mastery of the knowledge points obtained through training.
[0014] Furthermore, the step 1 is specifically as follows:
[0015] Given a set of exercises Gather with students There are N exercises and M students, using the student-exercise answer matrix Store all students' answer records, where X nm =1 means student St m Correctly answer the Exercise Ex n If the answer is wrong, then X nm =0;
[0016] Given K knowledge points, construct an exercise-knowledge point association matrix And Q nk =1 means Exercise Ex n Contains knowledge points Kc k, otherwise it is not included.
[0017] Furthermore, the step 2 is specifically as follows:
[0018] According to the student-exercise answer matrix X and the exercise-knowledge point association matrix Q, based on the non-negative matrix co-factorization technology, an encoding module with the following optimization objective function is constructed to represent students, exercises, and knowledge points in a low-dimensional latent space:
[0019]
[0020] in, represents the Frobenius norm; ⊙ represents the Hadamard product, which allows two matrices with the same dimension to produce another matrix with the same dimension, where each element of the new matrix is the product of the elements at the same position of the original two matrices. as well as They represent the exercise feature matrix, student ability matrix, and knowledge point requirement matrix in the low-dimensional latent space, respectively, where T represents the dimension of the (customized) low-dimensional latent space, and each low-dimensional matrix regards each dimension in its low-dimensional direction as a subject skill;
[0021] For the above encoding module: First, based on the Frobenius norm, the inner product of the exercise feature matrix E and the student ability matrix U is used as a low-rank approximation of X; since there are missing data in the matrix X, the weight matrix is introduced To indicate that the model only focuses on the numerical part of X. In W, the value of each element is either 0 or 1. nm = 1 if and only if X nm There is a record of answering questions; secondly, introduce the exercise-knowledge point correlation matrix Among them B nk Represents knowledge point Kc k For Exercise Ex n The importance of the question is expressed in terms of the exercise feature matrix E and the knowledge point requirement matrix V. The Frobenius norm is used to make a low-rank approximation of the matrix B, and the Q matrix is used as a sparse constraint. Finally, the constraint terms consider the non-negativity requirements for all variables to ensure that the better the students master the knowledge points, the more helpful it is to correctly answer related exercises.
[0022] Furthermore, the step three is specifically as follows:
[0023] Defining the Student-Knowledge Point Proficiency Matrix Among them, the non-negative elements A km Indicates student St m For knowledge point Kck Specific mastery of the problem, and introduce the difficulty vector of the problem The difficulty value of the simulation exercise, where μ n Representation Exercises n The difficulty coefficient of the problem is: based on the exercise-knowledge point association matrix B, the student-knowledge point proficiency matrix A, and the exercise difficulty vector M, a decoding module with the following optimization objective function is constructed to represent the student's mastery of the knowledge point:
[0024]
[0025] γ B represents the weight factor that balances the objective function (the first term) and the regularization term (the second term), represents the L2-norm. For the first term in the above formula, l is the log-likelihood function with cross entropy loss, which is defined as follows:
[0026]
[0027] in Contains the index position of the non-missing data in matrix X; B n: and A :m Respectively represent Exercise Ex n The knowledge point association vector and student St m The knowledge point proficiency vector of ; Φ(x) represents the standardized inverse link function, which is used for nonlinear mapping to simulate the student's answering process:
[0028]
[0029] here, is the probability density function with a standard normal distribution.
[0030] Furthermore, the construction of the student cognitive model can be transformed into solving an optimization problem with the following constraints:
[0031]
[0032] In the formula, O AF Represents the objective function of the student cognitive model.
[0033] Furthermore, the specific steps of model training in step 5 are as follows:
[0034] (a) Initialize the exercise feature matrix E, the student ability matrix U, the knowledge point requirement matrix V, and the exercise difficulty vector M;
[0035] (b) Calculate the objective function
[0036] (c) Set the external maximum number of iterations;
[0037] (d) For each round of external iteration, update the exercise-knowledge point association block matrix B in sequence n: (1≤n≤N), student ability block matrix U :m (1≤m≤M), exercise feature matrix E, knowledge point requirement matrix V, and exercise difficulty vector M;
[0038] (e) Calculate the new objective function again and compare the difference with the objective function of the previous round. If it is less than the preset threshold, stop the iteration and output all parameters. Otherwise, continue to execute (d).
[0039] Furthermore, step (d) is performed according to parameter B n: , U :m , E, V, and μ n ∈M, the target subfunction, gradient information and Lipschitz constant, set the exercise-knowledge point association block matrix B n: , student ability block matrix U :m , exercise feature matrix E, knowledge point requirement matrix V, and exercise difficulty value μ n The iterative update formula of ∈M is as follows:
[0040]
[0041]
[0042] Where [x]+=max(∈,x),∈=10 -15 To ensure the non-negativity of the output result; the superscripts (l) and (l+1) represent the (l)th and (l+1)th iterations respectively, and L(·) represents the Lipschitz constant. Represents gradient information.
[0043] Furthermore, the parameter B n: , U :m , E, V, and μ n The specific calculation of the target subfunction, gradient information and Lipschitz constant is:
[0044] (i) Target sub-function:
[0045]
[0046] in,
[0047] (ii) Gradient information:
[0048]
[0049]
[0050] in
[0051] (iii) Lipschitz constant:
[0052]
[0053] Among them, L p Representation function The Lipschitz constant, whose value can be derived as 1; Φ'(x) represents the first-order derivative of Φ(x); σ1(·) represents the maximum value of its elements; J M,N Represents a matrix of dimension M×N with all elements set to 1.
[0054] Furthermore, the step six is specifically as follows: fitting the students' answer results through the training objective function, calculating the exercise feature matrix E, the student ability matrix U, and the student-knowledge point proficiency matrix A, using E and U to complete the learning performance prediction task, and using A to complete the knowledge level diagnosis task.
[0055] Compared with the prior art, the present invention has the following beneficial effects:
[0056] (a) At the knowledge level, the present invention makes full use of students' answer records to exercises, regards them as the input of the encoding module and the output of the decoding module, and introduces the student-knowledge point proficiency matrix to explicitly represent the students' knowledge level, breaking the barrier that traditional educational data mining methods cannot effectively simulate students' knowledge level.
[0057] (b) At the problem level, considering that the Q matrix only roughly represents the relationship between problems and knowledge points, and does not reflect the strength of the relationship between the two, the latter is not only very important for students' cognitive analysis, but also plays a key guiding role in helping students choose appropriate problems in personalized recommendation tasks in the field of education. Therefore, the present invention introduces the problem-knowledge point correlation matrix to refine the correlation strength of different knowledge points to any problem.
[0058] (c) At the solution level, the present invention introduces an inverse link function to perform nonlinear mapping between the knowledge feature vector and the exercise feature vector in the low-dimensional space to simulate the student's answering process, so as to better adapt to the complexity of the student's cognitive process, including the nonlinear changes in the learning curve and the stages of concept mastery. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] Figure 1 A flow chart of a student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization provided by an embodiment of the present invention;
[0060] Figure 2 Provide students with cognitive modeling assessment scenario diagrams;
[0061] Figure 3 A schematic diagram of a student cognitive model based on autoencoder-like non-negative matrix co-factorization provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0062] The following is combined with Figure 1-3 , the technical solution of the present invention is specifically described.
[0063] The present invention proposes a student cognitive model based on autoencoder-like non-negative matrix co-factorization, such as Figure 1 , 3 As shown, it mainly includes the following steps:
[0064] (1) Step 1 (data acquisition): Obtain students’ historical answer data (correct answers, wrong answers, or scores) and exercise-knowledge point association information (Q matrix). Given an exercise set Gather with students (There are N exercises and M students), using the student-exercise answer matrix Store all students' answer records, where X nm =1 means student St m Correctly answer the Exercise Ex n If the answer is wrong, then X nm = 0. In addition, given K knowledge points, construct the exercise-knowledge point association matrix And Q nk =1 means Exercise Ex n Contains knowledge points Kc k ; Otherwise it is not included.
[0065] (2) Step 2 (Encoding Module): Based on the student-exercise answer matrix X and the exercise-knowledge point association matrix Q, an encoding module with the following optimization objective function is constructed to represent students, exercises, and knowledge points in a low-dimensional latent space based on the non-negative matrix co-factorization technology:
[0066]
[0067] in, represents the Frobenius norm, or Frobenius norm for short; ⊙ represents the Hadamard product, which allows two matrices with the same dimension to produce another matrix with the same dimension, where each element of the new matrix is the product of the elements in the same position of the original two matrices, as well as They represent the exercise feature matrix, student ability matrix, and knowledge point requirement matrix in the low-dimensional latent space respectively. Where T represents the dimension of the (customized) low-dimensional latent space, and for each low-dimensional matrix, each dimension in the low-dimensional direction is regarded as a subject skill.
[0068] For the above encoding module: First, based on the Frobenius norm, the inner product of the exercise feature matrix E and the student ability matrix U is used as a low-rank approximation of X; since there are missing data in the matrix X, the weight matrix is introduced To indicate that the model only focuses on the numerical part of X. In W, the value of each element is either 0 or 1. nm = 1 if and only if X nm There is a record of answering questions; secondly, introduce the exercise-knowledge point correlation matrix Among them B nk Represents knowledge point Kc k For Exercise Ex n The importance of , and the larger the value, the deeper the degree of association. The exercise feature matrix E and the knowledge point requirement matrix V are used to make a low-rank approximation of the matrix B, and the Q matrix is used as a sparse constraint. Finally, the constraint term considers the non-negativity requirement for all variables to ensure that the better the students master the knowledge points, the more helpful it is to correctly answer related exercises.
[0069] (3) Step 3 (Decoding Module): Define the Student-Knowledge Point Proficiency Matrix Among them, the non-negative elements A km Indicates student St m For knowledge point Kc k Specific mastery of the problem, and introduce the difficulty vector M = The difficulty value of the simulation exercise, where μ n Representation Exercises n Based on the exercise-knowledge point association matrix B, the student-knowledge point proficiency matrix A, and the exercise difficulty vector M, a decoding module with the following optimization objective function is constructed to represent the student's mastery of the knowledge point:
[0070]
[0071] γ B represents the weight factor that balances the objective function (the first term) and the regularization term (the second term), represents the L2-norm. For the first term in the above formula, l is the log-likelihood function with cross entropy loss, which is defined as follows:
[0072]
[0073] in Contains the index positions of non-missing data in matrix X. Bn: and A :m Respectively represent Exercise Ex n The knowledge point association vector and student St n The knowledge point proficiency vector of . Φ(x) represents the standardized inverse link function, which is used for nonlinear mapping to simulate the student's answering process:
[0074]
[0075] here, is the probability density function with a standard normal distribution.
[0076] (4) Step 4 (module fusion): After integrating the encoding module and the decoding module, the constructed student cognitive model can be transformed into solving an optimization problem with the following constraints:
[0077]
[0078] In the formula, Represents the objective function of the student cognitive model.
[0079] (5) Step 5 (model training): Consider: the model to be solved is not a joint convex function of the parameters E, U, and V, and has non-negative constraints; the exercise-knowledge point association matrix B has a block attribute, so the present invention adopts a constrained block coordinate descent projection gradient method. At the same time, considering the impact of iteration speed, the present invention proposes a step-size solution strategy based on the Lipschitz constant. Specifically, regarding the parameter B n: , U :m , E, V, and μ n The objective subfunction, gradient information and Lipschitz constant are as follows:
[0080] (i) Target sub-function:
[0081]
[0082] in,
[0083] (ii) Gradient information:
[0084]
[0085] in
[0086] (iii) Lipschitz constant:
[0087]
[0088] Among them, L p Representation function The Lipschitz constant, whose value can be derived as 1; Φ'(x) represents the first-order derivative of Φ(x); σ1(·) represents the maximum value of its elements; J M,N Represents a matrix with dimension M×N and all element values are 1;
[0089] Based on the gradient information and the Lipschitz constant, B n: For example, the iterative update formula can be described as follows (the other parameters are similar):
[0090]
[0091] Here [x] + =max(∈,x),∈=10 -15 To ensure the non-negativity of the output result, the superscripts (l) and (l+1) represent the (l)th and (l+1)th iterations respectively, and L(·) represents the Lipschitz constant. Represents gradient information.
[0092] Specifically, the complete iterative process is:
[0093] (a) Initialize the exercise feature matrix E, the student ability matrix U, the knowledge point requirement matrix V, and the exercise difficulty vector M;
[0094] (b) Calculate the objective function
[0095] (c) Set the external maximum number of iterations;
[0096] (d) For each round of external iteration, update the exercise-knowledge point association block matrix B in sequence n: (1≤n≤N), student ability block matrix U :m (1≤m≤M), exercise feature matrix E, knowledge point requirement matrix V, and exercise difficulty vector M;
[0097] (e) Calculate the new objective function again and compare the difference with the objective function of the previous round. If it is less than the preset threshold, stop the iteration and output all parameters. Otherwise, continue to execute (d).
[0098] (6) Step 6 (cognitive analysis): Fit the students’ answers by training the objective function to obtain the exercise feature matrix E, the student ability matrix U, and the student-knowledge point proficiency matrix A. Use E and U to complete the learning performance prediction task, and use A to complete the knowledge level diagnosis task.
[0099] The above are preferred embodiments of the present invention. Any changes made according to the technical solution of the present invention, as long as the resulting functions do not exceed the scope of the technical solution of the present invention, belong to the protection scope of the present invention.
Claims
1. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization, characterized in that: The following steps are involved: Step 1: Obtain students’ historical answer data and extract the question-knowledge point association information in the test text; Step 2: Based on the students’ answer data and the relationship between exercises and knowledge points, a coding module is built to construct a low-dimensional latent space representation of students, exercises, and knowledge points; Step 3: Use the low-dimensional latent space information of students, exercises and knowledge points to build a decoding module to indicate the students' mastery of knowledge points; Step 4: Integrate the encoding module and the decoding module to complete the construction of the student cognitive model; Step 5: Use the constrained block coordinate descent projected gradient method and combine it with the Lipschitz step size solution strategy to train the model; Step 6: Complete the learning performance prediction task based on the exercises and the low-dimensional latent space representation of the students, and complete the knowledge level diagnosis task based on the students' mastery of the knowledge points obtained through training.
2. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 1, characterized in that: The step 1 is specifically as follows: Given a set of exercises Gather with students There are N exercises and M students, using the student-exercise answer matrix Store all students' answer records, where X nm =1 means student St m Correctly answer the Exercise Ex n If the answer is wrong, then X nm =0; Given K knowledge points, construct the exercise-knowledge point association matrix And Q nk =1 means Exercise Ex n Contains knowledge points Kc k , otherwise it is not included.
3. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 2, characterized in that: The step 2 is specifically as follows: According to the student-exercise answer matrix X and the exercise-knowledge point association matrix Q, based on the non-negative matrix co-factorization technology, an encoding module with the following optimization objective function is constructed to represent students, exercises, and knowledge points in a low-dimensional latent space: in, represents the Frobenius norm; ⊙ represents the Hadamard product, which allows two matrices with the same dimension to produce another matrix with the same dimension, and each element of the new matrix is the product of the elements in the same position of the original two matrices; as well as Respectively represent the exercise feature matrix, student ability matrix, and knowledge point requirement matrix in the low-dimensional latent space, T represents the dimension of the (customized) low-dimensional latent space, and each low-dimensional matrix regards each dimension in its low-dimensional direction as a subject skill; For the above encoding module: First, based on the Frobenius norm, the inner product of the exercise feature matrix E and the student ability matrix U is used as a low-rank approximation of X. Since there are missing data in the matrix X, the weight matrix is introduced To indicate that the model only focuses on the numerical part of X, in W, the value of each element is either 0 or 1, W nm = 1 if and only if X nm There is a record of answering questions; secondly, introduce the exercise-knowledge point correlation matrix Among them B nk Represents knowledge point Kc k For Exercise Ex n The importance of the question is expressed in terms of the exercise feature matrix E and the knowledge point requirement matrix V. The Frobenius norm is used to make a low-rank approximation of the matrix B, and the Q matrix is used as a sparse constraint. Finally, the constraint terms consider the non-negativity requirements for all variables to ensure that the better the students master the knowledge points, the more helpful it is to correctly answer related exercises.
4. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 3, characterized in that: The step three is specifically as follows: Defining the Student-Knowledge Point Proficiency Matrix Among them, the non-negative elements A km Indicates student St m For knowledge point Kc k Specific mastery of the problem, and introduce the difficulty vector of the problem The difficulty value of the simulation exercise, where μ n Representation Exercises n The difficulty coefficient of the problem is: based on the exercise-knowledge point association matrix B, the student-knowledge point proficiency matrix A, and the exercise difficulty vector M, a decoding module with the following optimization objective function is constructed to represent the student's mastery of the knowledge point: γ B represents the weight factor that balances the objective function and the regularization term, represents the L2-norm. For the first term in the above formula, l is the log-likelihood function with cross entropy loss, which is defined as follows: in Contains the index position of the non-missing data in matrix x; B n: and A :m Respectively represent Exercise Ex n The knowledge point association vector and student St m The knowledge point proficiency vector of ; Φ(x) represents the standardized inverse link function, which is used for nonlinear mapping to simulate the student's answering process: here, is the probability density function with a standard normal distribution.
5. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 4, characterized in that: The construction of the student cognitive model can be transformed into solving an optimization problem with the following constraints: stB≥0,E≥0,U≥0,V≥0. In the formula, Represents the objective function of the student cognitive model.
6. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 5, characterized in that: The specific steps of model training in step 5 are as follows: (a) Initialize the exercise feature matrix E, the student ability matrix U, the knowledge point requirement matrix V, and the exercise difficulty vector M; (b) Calculate the objective function (c) Set the external maximum number of iterations; (d) For each round of external iteration, update the exercise-knowledge point association block matrix B in sequence n :(1≤n≤N), student ability block matrix U:m(1≤m≤M), exercise feature matrix E, knowledge point requirement matrix V, and exercise difficulty vector M; (e) Calculate the new objective function again and compare the difference with the objective function of the previous round. If it is less than the preset threshold, stop the iteration and output all parameters. Otherwise, continue to execute (d).
7. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 6, characterized in that: Step (d) is based on parameter B n: , U :m , E, V, and μ n ∈M, the target subfunction, gradient information and Lipschitz constant, set the exercise-knowledge point association block matrix B n: , student ability block matrix U :m , exercise feature matrix E, knowledge point requirement matrix V, and exercise difficulty value μ n The iterative update formula of ∈M is as follows: Where [x] + =max(∈, x),∈=10 -15 To ensure the non-negativity of the output result; the superscripts (l) and (l+1) represent the (l)th and (l+1)th iterations respectively, and L(·) represents the Lipschitz constant. Represents gradient information.
8. A student cognitive modeling method based on autoencoder-like non-negative matrix co-factorization according to claim 7, characterized in that: The parameter B n: , U :m , E, V, and μ n The specific calculation of the target subfunction, gradient information and Lipschitz constant is: (i) Target sub-function: in, (ii) Gradient information: in (iii) Lipschitz constant: Among them, L p Representation function The Lipschitz constant, whose value can be derived as 1; Φ′(x) represents the first-order derivative of Φ(x); σ1(·) represents the maximum value of its elements; J M,N Represents a matrix of dimension M×N with all elements set to 1.
9. The method for student cognitive modeling based on autoencoder-like non-negative matrix co-factorization according to claim 6, characterized in that: The step six is specifically as follows: fit the students' answer results through the training objective function, calculate the exercise feature matrix E, the student ability matrix U, and the student-knowledge point proficiency matrix A, use E and U to complete the learning performance prediction task, and use A to complete the knowledge level diagnosis task.
Citation Information
Patent Citations
Student cognitive diagnosis method
CN110264091A
Student performance prediction method combining global and local features
CN116166998A
Sparse Factor Analysis for Learning Analytics and Content Analytics
US20140272914A1
