CN116306863A - Collaborative knowledge tracking modeling method and system based on contrast learning - Google Patents
Collaborative knowledge tracking modeling method and system based on contrast learning Download PDFInfo
- Publication number
- CN116306863A CN116306863A CN202310019414.8A CN202310019414A CN116306863A CN 116306863 A CN116306863 A CN 116306863A CN 202310019414 A CN202310019414 A CN 202310019414A CN 116306863 A CN116306863 A CN 116306863A
- Authority
- CN
- China
- Prior art keywords
- students
- knowledge
- similar
- collaborative
- answer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 230000006870 function Effects 0.000 claims abstract description 14
- 230000006399 behavior Effects 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 5
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000003993 interaction Effects 0.000 description 7
- 238000012360 testing method Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Educational Technology (AREA)
- Educational Administration (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The collaborative knowledge tracking modeling method based on contrast learning comprises collaborative prediction: before each question is made, the students determine the similarity between the historical answer sequences of the students and the historical answer sequences of other students by calculating the intersection ratio, finally, K most similar students are searched, the state of common knowledge in the students and the answer performance at the next moment are extracted, and the historical answer sequences of the original students are cooperated to perform corresponding prediction; contrast study: in the retrieved similar students, the knowledge states are similar when the same item is made, namely, a pair of positive examples is formed by the knowledge states of the similar students and the knowledge states of the original students at the moment, and the knowledge states of the similar students at all other moments and the knowledge states of the original students at the moment form negative examples, so that the representation of each knowledge state is more accurate after the loss function is compared; the invention has the beneficial effects that: the teaching machine can improve the teaching quality and realize personalized teaching.
Description
Technical Field
The invention belongs to the field of intelligent education and education big data mining, and particularly relates to a collaborative knowledge tracking modeling method and system based on contrast learning.
Background
Online education is evolving on an unprecedented scale.
Meanwhile, according to the actual situation of each student, a learning plan is formulated, educational resources can be efficiently utilized, learning efficiency and educational output are improved, and the educational means of "teaching in accordance with the material" is deeply concerned by educational workers. In the conventional educational outcome evaluation mechanism, merely evaluating the knowledge mastery level of a student by means of the final score of an examination cannot embody the mastery condition of each knowledge concept by the student. For example, even if the first and second students obtain the same score in the same test question, the two students cannot simply consider that the level of knowledge mastered by the two students is the same, because there may be a case where the student first answers the wrong test question and the student second answers the right. Even if two students make mistakes in the same topic, it is possible that two students make mistakes at different knowledge points examined in the topic. The modeling of the mastering level of students on each knowledge concept is a basic premise for realizing teaching based on the material, implementing personalized teaching and improving the teaching working quality.
Knowledge tracking (Knowledge Tracing, KT) models knowledge mastering conditions of students according to past answering conditions of the students, so that the knowledge mastering degree of the students on knowledge points can be accurately predicted. The knowledge tracking problem may be formed as a group of students S and a group of exercises E in one student answer interaction, where different students are required to answer different exercises to reach a grasp of the relevant knowledge. Each exercise is associated with a specific knowledge concept. The learning order of the student is denoted i= (e_1, a_1), (e_2, a_2), …, (e_n, a_n), where e_t represents the t-th exercise the student answers, a_t represents an error-positive tag (i.e. 1 represents a correct answer, 0 represents a wrong answer), and N represents the sequence length of the learning interaction. The research content of knowledge tracking is to model the answer sequence of the student to obtain the knowledge state of the student (the mastering degree of the student on each knowledge point) at each moment, so that the probability of answering the question by the student can be predicted according to the knowledge state of the student before the student makes the next question.
Most of methods based on deep neural networks only use the student's own historical answer sequences to predict and use simple sequence models to represent knowledge states at each moment. This approach is too simple, does not fully exploit the information, and does not distinguish well between representations of knowledge states, which greatly limits further improvement of model performance. In a real environment, when a teacher judges whether a student can answer a current question, the teacher can refer to the historical answer performance of the student, meanwhile, the performance of similar students on the question can be considered, and multiple information is synthesized to perform corresponding judgment.
Disclosure of Invention
In order to solve the defects in the prior art, improve the teaching work quality and realize personalized teaching, the invention discloses a collaborative knowledge tracking modeling method and system based on contrast learning; according to the invention, the sequence modeling is performed by using the historical answer sequences of the students and the historical answer sequences of similar students, the obtained two knowledge state representations and the question making representation of the similar students on the problem are fused, and the representation of the further accurate knowledge state is used for comparison learning, so that the comprehensive representation is finally obtained to predict the answer representation of the students, and the accuracy of model prediction is improved.
The application provides a collaborative knowledge tracking modeling method based on contrast learning, which comprises collaborative prediction: before each question is made, the students search all the students who have made the question or the concept, the similarity between the historical answer sequences of the students and the historical answer sequences of the original students is determined through calculating the intersection ratio, finally K most similar students are searched, the common knowledge states of the students and the answer expression at the next moment are extracted, and the historical answer sequences of the original students are cooperated to perform corresponding prediction;
contrast study: in the retrieved similar students, the knowledge states are similar when the same item is made, namely, a pair of positive examples is formed by the knowledge states of the similar students and the knowledge states of the original students at the moment, and the knowledge states of the similar students at all other moments and the knowledge states of the original students at the moment form negative examples, so that the representation of each knowledge state is more accurate after the loss function is compared;
the method also comprises the following steps:
(1) Problem definition and data preprocessing;
(2) Retrieving similar students;
(3) Modeling a gating unit LSTM;
(4) Contrast learning;
(5) Collaborative prediction, loss calculation and optimization functions.
Preferably, in the problem definition and the data preprocessing, in the two data sets, namely, the ASSIST2009 and the ASSIST2017, the actual student answer sequences often have the situations of short answer sequences (the number of the answers is less than 2), blank values (the questions lack knowledge points or the knowledge points do not have corresponding questions), and the like. For shorter sequences of student sample data, and for information with incomplete data, removal is required during the preprocessing stage. For the EdNet dataset, since the original data is too huge (about 78 thousands of students data), 5000 students were randomly selected from the original data, and then the processing manner is the same as the two datasets. In order to intuitively show the answer sequences of students, the answer sequence of each student in the data set is represented by four rows: the number of answers of the first behavior answer sequence; question numbers of the second behavior answer sequence; knowledge points corresponding to the third behavior question numbers; and fourth, student answer pair or not (1 is answer pair, 0 is answer mistake).
Preferably, the similar students are searched, namely before each question is made by the students, all the students who have made the question or the concept are searched, the similarity between the historical answer sequences of the students and the historical answer sequences of the original students is determined by calculating the intersection ratio, and finally the K most similar students are searched.
Preferably, the gating unit LSTM modeling includes:
h t =o t *tanh(C t )
preferably, the collaborative prediction, loss calculation and optimization function includes:
(1) Obtaining knowledge state h of original student at time t in LSTM t Knowledge state c of the same student before doing the same question t And the performance r of similar students on the current questions t+1 And splice the threeObtaining v t . Next, v t And weight matrix W yv Bias b y Calculating to obtain y after sigmoid function sigma (& gt) t+1 ,y t+1 Is a matrix containing N number of questions, the probability of each question corresponding to a pair (between 0 and 1):
y t+1 =σ(W yv v t +b y )
will y t+1 A thinking answer to the underlying question (i.e., q) with a predictive value of 0.5 or more t+1 =1), consider the solution to be wrong (q t+1 =0). And true (q t+1 ,a t+1 ) Comparing, calculating loss
(2) In a learning sequence of N students in a minimatch, there are n×l knowledge states, and then there are n×l similar knowledge states correspondingly, we will know knowledge state h of a student at time t t And similar knowledge state c at this time t As a pair of positive examples, and the remaining 2 (N. Times.L-1) knowledge states as negative examples, a contrast loss was calculated
The collaborative knowledge tracking modeling system based on contrast learning comprises a terminal device and a server, wherein the server comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, and the processor realizes all the methods when executing the program.
Compared with the prior art, the invention has the following beneficial effects:
1. by searching out similar students to conduct collaborative prediction, compared with the traditional method for predicting by only using own historical answer sequences, the method provided by the invention has the advantage that the related information of the similar students is added, so that the knowledge state of the students can be expressed more accurately.
2. The method can obtain a better knowledge state representation by utilizing the relationship between the knowledge states of similar students compared with the traditional method of modeling by using a simple sequence model only. Compared with the existing method, the accuracy of the knowledge tracking model can be remarkably improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments or the description of the prior art will be briefly described below, it will be obvious that the drawings in the following description are some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort to a person skilled in the art.
Fig. 1 is an AUC index value for each of the methods of the present invention.
FIG. 2 is a schematic representation of the process of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiment one: collaborative knowledge tracking modeling method based on contrast learning
The specific implementation steps are as follows:
(1) Problem definition and data preprocessing
In the prior art, no matter how to optimize a sequence model and add auxiliary information, similar student information is not considered for collaborative prediction, but actually, in a real environment, when a teacher judges whether a student can answer a current question, the teacher can refer to the historical answer performance of the student, meanwhile, the performance of the similar student is considered, and multiple kinds of information are synthesized for corresponding judgment.
The invention is mainly used for experiments on three published data sets commonly used for knowledge tracking problems. In two data sets, namely an ASSIST2009 and an ASSIST2017, the actual student answer sequences often have the situations of short answer sequences (the number of the answers is less than 2), blank values (the test questions lack of knowledge points or the knowledge points do not have corresponding test questions) and the like. For shorter sequences of student sample data, and for information with incomplete data, removal is required during the preprocessing stage. For the EdNet dataset, since the original data is too huge (about 78 thousands of students data), 5000 students were randomly selected from the original data, and then the processing manner is the same as the two datasets. In order to intuitively show the answer sequences of students, the answer sequence of each student in the data set is represented by four rows: the first line is the number of answers of the answer sequence; the second line is the question number of the answer sequence; the third row is the knowledge point corresponding to the question number; the fourth line is the student answer pair or not (1 is answer pair, 0 is answer error).
(2) Retrieving similar students
Interaction at time i for any one student uHistorical sequenceWe retrieve all records from the dataset that meet the following conditions as candidate records
(2) The two interactions have the same concept or problem, i.e. c i =c j or e i =e j
All candidate recorded students are similar students, and their history sequence is To further obtain the K students that have the most similar information at student u's time i, we use the intersection ratio to calculate the similarity between the current student and the candidate students, and select the K students with the largest intersection ratio as similar students.
(3) Gated loop unit LSTM modeling
The method solves the problem that the number of answers of each student is different, the step length is set to be 50 (less than 50 is filled and supplemented by zero), each 50 questions are in a group, each interaction record is initialized and expressed in the following mode, and then the interaction records are input into a gating circulation unit LSTM, and the obtained output is mapped into N-dimensional representation to obtain the probability of each question answer pair.
h t =o t *tanh(C t )
(4) Contrast learning
In a learning sequence of N students in a minimatch, there are n×l knowledge states, and correspondingly n×l similar knowledge states exist, we will know knowledge state h at time t of a student t And similar knowledge state c at this time t As a pair of positive examples, the remaining 2 (N x L-1) knowledge states are taken as negative examples.
sim(u,v)=u T v/|u||v|
(5) Collaborative prediction, loss calculation and optimization function
Obtaining knowledge state h of original student at time t in LSTM t Knowledge state c of the same student before doing the same question t Representation r of similar students on the current question t+1 And splicing the three to obtain v t . Next, v t And weight matrix W yv Bias b y Calculating to obtain y after sigmoid function sigma (& gt) t+1 ,y t+1 Is a matrix containing N number of questions, the probability of each question corresponding to a pair (between 0 and 1):
y t+1 =σ(W yv v t +b y )
will y t+1 A thinking answer to the underlying question (i.e., q) with a predictive value of 0.5 or more t+1 =1), consider the solution to be wrong (q t+1 =0). And true (q t+1 ,a t+1 ) Comparing, calculating loss
Simulation verification is carried out on the implementation method, and fig. 1 shows the performance of the method on the public data sets ASSIST2009, ASSIST2017 and EdNet, and compared with the existing 4 depth knowledge tracking modeling methods (respectively marked as DKT, DKVMN, SAKT, AKT), the method has better performance. The invention utilizes AUC index to measure the performance of each method, wherein AUC (Area Under Curve) is a common index for measuring the performance of a model method, and the meaning of the AUC index is the area of an area surrounded by an ROC curve and a coordinate axis.
Embodiment two: collaborative knowledge tracking modeling system based on contrast learning
In one or more embodiments, a terminal device is disclosed that includes a server including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the two-channel graph neural network-based knowledge tracking modeling method of embodiment one when executing the program. For brevity, the description is omitted here. It should be understood that the processor in this embodiment may be a central processing unit CPU, and the processor may also be other general purpose processors, digital signal processors DSP, application specific integrated circuits ASIC, off-the-shelf programmable gate array FPGA or other programmable logic device, discrete gate or transistor logic devices, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The memory may include read only memory and random access memory and provide instructions and data to the processor, and a portion of the memory may also include non-volatile random access memory. For example, the memory may also store information of the device type. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or by instructions in the form of software. The knowledge tracking modeling method based on the dual-channel graph neural network in the first embodiment can be directly embodied as the execution completion of a hardware processor or the execution completion of the combination execution of hardware and software modules in the processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory, and the processor reads the information in the memory and, in combination with its hardware, performs the steps of the above method. To avoid repetition, a detailed description is not provided herein.
Finally, it should be noted that: those of ordinary skill in the art will appreciate that the elements of the various examples described in connection with the present embodiments, i.e., the algorithm steps, can be implemented as electronic hardware or combinations of computer software and electronic hardware; whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution; skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application; while the foregoing description of the embodiments of the present invention has been presented in conjunction with the drawings, it should be understood that it is not intended to limit the scope of the invention, but rather, it is intended to cover all modifications or variations within the scope of the invention as defined by the claims of the present invention.
Claims (6)
1. The collaborative knowledge tracking modeling method based on contrast learning is characterized by comprising the following steps of: including collaborative prediction: before each question is made, the students search all the students who have made the question or the concept, the similarity between the historical answer sequences of the students and the historical answer sequences of the original students is determined through calculating the intersection ratio, finally K most similar students are searched, the common knowledge states of the students and the answer expression at the next moment are extracted, and the historical answer sequences of the original students are cooperated to perform corresponding prediction;
contrast study: the knowledge states of similar students are similar when doing the same item, namely, the knowledge states of the similar students and the knowledge states of the original students at the moment form a pair of positive examples, the knowledge states of the corresponding other moment and the knowledge states of the original students at the moment form negative examples, and the representation of each knowledge state is more accurate after the loss function is compared;
the method also comprises the following steps:
(1) Problem definition and data preprocessing;
(2) Retrieving similar students;
(3) Modeling a gating unit LSTM;
(4) Contrast learning;
(5) Collaborative prediction, loss calculation and optimization functions.
2. The collaborative knowledge tracking modeling method based on contrast learning according to claim 1, wherein: the problem definition and the data preprocessing are that in two data sets, namely an ASSIST2009 and an ASSIST2017, in order to intuitively embody the answer sequences of students, the answer sequences of each student in the data sets are represented by four rows: the number of answers of the first behavior answer sequence; question numbers of the second behavior answer sequence; knowledge points corresponding to the third behavior question numbers; fourth behavior students answer or not.
3. The collaborative knowledge tracking modeling method based on contrast learning according to claim 1, wherein: before the students do each question, the similar students are searched, all the students doing the question or the concept are searched, the similarity between the historical answer sequences of the students and the historical answer sequences of the original students is determined through calculating the intersection ratio, and finally the K most similar students are searched.
5. the collaborative knowledge tracking modeling method based on contrast learning according to claim 1, wherein: the collaborative prediction, loss calculation and optimization function includes:
(1) Obtaining knowledge state h of original student at time t in LSTM t Knowledge state c of the same student before doing the same question t And the performance r of similar students on the current questions t+1 And splicing the three to obtain v t . Next, v t And weight matrix W yv Bias b y Calculating to obtain y after sigmoid function sigma (& gt) t+1 ,y t+1 Is a matrix containing N number of questions, the probability of each question corresponding to a pair (between 0 and 1):
y t+1 =σ(W yv v t +b y )
will y t+1 A thinking answer to the underlying question (i.e., q) with a predictive value of 0.5 or more t+1 =1), consider the solution to be wrong (q t+1 =0). And true (q t+1 ,a t+1 ) Comparing, calculating loss
(2) N students in a minibandIn the learning sequence of (a), there are n×l knowledge states, so there are n×l similar knowledge states correspondingly, we will be a knowledge state h of student at time t t And similar knowledge state c at this time t As a pair of positive examples, and the remaining 2 (N. Times.L-1) knowledge states as negative examples, a contrast loss is calculated
6. Collaborative knowledge tracking modeling system based on contrast learning, including a terminal equipment, its characterized in that: the collaborative knowledge tracking modeling method based on the contrast learning comprises the server, wherein the server comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, and the processor realizes the collaborative knowledge tracking modeling method based on the contrast learning when executing the program.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310019414.8A CN116306863A (en) | 2023-01-06 | 2023-01-06 | Collaborative knowledge tracking modeling method and system based on contrast learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310019414.8A CN116306863A (en) | 2023-01-06 | 2023-01-06 | Collaborative knowledge tracking modeling method and system based on contrast learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116306863A true CN116306863A (en) | 2023-06-23 |
Family
ID=86829460
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310019414.8A Pending CN116306863A (en) | 2023-01-06 | 2023-01-06 | Collaborative knowledge tracking modeling method and system based on contrast learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116306863A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117077737A (en) * | 2023-08-22 | 2023-11-17 | 长江大学 | Knowledge tracking system for dynamic collaboration of knowledge points |
CN117744783A (en) * | 2024-01-29 | 2024-03-22 | 暨南大学 | Knowledge tracking method and system based on man-in-the-loop |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112949929A (en) * | 2021-03-15 | 2021-06-11 | 华中师范大学 | Knowledge tracking method and system based on collaborative embedded enhanced topic representation |
CN113033808A (en) * | 2021-03-08 | 2021-06-25 | 西北大学 | Deep embedded knowledge tracking method based on exercise difficulty and student ability |
CN114547267A (en) * | 2022-02-22 | 2022-05-27 | 武汉纺织大学 | Intelligent question-answering model generation method and device, computing equipment and storage medium |
CN115328971A (en) * | 2022-08-23 | 2022-11-11 | 山东财经大学 | Knowledge tracking modeling method and system based on double-graph neural network |
CN115329096A (en) * | 2022-08-19 | 2022-11-11 | 上海伯禹教育科技有限公司 | Interactive knowledge tracking method based on graph neural network |
CN115510286A (en) * | 2022-11-08 | 2022-12-23 | 合肥工业大学 | Multi-relation cognitive diagnosis method based on graph convolution network |
-
2023
- 2023-01-06 CN CN202310019414.8A patent/CN116306863A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113033808A (en) * | 2021-03-08 | 2021-06-25 | 西北大学 | Deep embedded knowledge tracking method based on exercise difficulty and student ability |
CN112949929A (en) * | 2021-03-15 | 2021-06-11 | 华中师范大学 | Knowledge tracking method and system based on collaborative embedded enhanced topic representation |
CN114547267A (en) * | 2022-02-22 | 2022-05-27 | 武汉纺织大学 | Intelligent question-answering model generation method and device, computing equipment and storage medium |
CN115329096A (en) * | 2022-08-19 | 2022-11-11 | 上海伯禹教育科技有限公司 | Interactive knowledge tracking method based on graph neural network |
CN115328971A (en) * | 2022-08-23 | 2022-11-11 | 山东财经大学 | Knowledge tracking modeling method and system based on double-graph neural network |
CN115510286A (en) * | 2022-11-08 | 2022-12-23 | 合肥工业大学 | Multi-relation cognitive diagnosis method based on graph convolution network |
Non-Patent Citations (1)
Title |
---|
TING LONG ET AL: "Improving Knowle dge Tracing with Collab orative Information", 《WSDM》, pages 599 - 603 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117077737A (en) * | 2023-08-22 | 2023-11-17 | 长江大学 | Knowledge tracking system for dynamic collaboration of knowledge points |
CN117077737B (en) * | 2023-08-22 | 2024-03-15 | 长江大学 | Knowledge tracking system for dynamic collaboration of knowledge points |
CN117744783A (en) * | 2024-01-29 | 2024-03-22 | 暨南大学 | Knowledge tracking method and system based on man-in-the-loop |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116306863A (en) | Collaborative knowledge tracking modeling method and system based on contrast learning | |
CN111582694A (en) | Learning evaluation method and device | |
CN114968788A (en) | Method, apparatus, medium, and device for automatically evaluating programming capability of artificial intelligence algorithm | |
Ding et al. | Why Deep Knowledge Tracing Has Less Depth than Anticipated. | |
CN114021722A (en) | Attention knowledge tracking method integrating cognitive portrayal | |
CN109272160A (en) | Score on Prediction system and prediction technique | |
CN114254127A (en) | Student ability portrayal method and learning resource recommendation method and device | |
CN114913729A (en) | Question selection method and device, computer equipment and storage medium | |
CN116186250A (en) | Multi-mode learning level mining method, system and medium under small sample condition | |
CN115328971A (en) | Knowledge tracking modeling method and system based on double-graph neural network | |
CN115310520A (en) | Multi-feature-fused depth knowledge tracking method and exercise recommendation method | |
CN114971066A (en) | Knowledge tracking method and system integrating forgetting factor and learning ability | |
CN113239699B (en) | Depth knowledge tracking method and system integrating multiple features | |
CN113283488B (en) | Learning behavior-based cognitive diagnosis method and system | |
WO2021147405A1 (en) | Customer-service statement quality detection method and related device | |
CN116361744A (en) | Learner cognition tracking method and system for learning procedural evaluation | |
CN115984053A (en) | Test question recommendation method and device, electronic equipment and storage medium | |
CN114117033B (en) | Knowledge tracking method and system | |
CN115422256A (en) | Interpretable knowledge cognition level mining method and system fusing learning cognition elements | |
CN115409257A (en) | Score distribution prediction method and system based on condition density estimation model | |
CN114741591A (en) | Method and electronic equipment for recommending learning path to learner | |
CN111339258A (en) | University computer basic exercise recommendation method based on knowledge graph | |
Li et al. | Evaluation method of GA-BP neural network programming ability based on entropy weight-deviation | |
Jebli et al. | Proposal of a similarity measure for unified modeling language class diagram images using convolutional neural network. | |
Yun et al. | [Retracted] Quality Evaluation and Satisfaction Analysis of Online Learning of College Students Based on Artificial Intelligence |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |