CN114723590A - Group-oriented knowledge tracking method, system, device and storage medium - Google Patents

Group-oriented knowledge tracking method, system, device and storage medium Download PDF

Info

Publication number
CN114723590A
CN114723590A CN202210323642.XA CN202210323642A CN114723590A CN 114723590 A CN114723590 A CN 114723590A CN 202210323642 A CN202210323642 A CN 202210323642A CN 114723590 A CN114723590 A CN 114723590A
Authority
CN
China
Prior art keywords
matrix
knowledge
feature
expression
dimension
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.)
Granted
Application number
CN202210323642.XA
Other languages
Chinese (zh)
Other versions
CN114723590B (en
Inventor
吴正洋
黄立
汤庸
陈展轩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
South China Normal University
Original Assignee
South China Normal University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by South China Normal University filed Critical South China Normal University
Priority to CN202210323642.XA priority Critical patent/CN114723590B/en
Publication of CN114723590A publication Critical patent/CN114723590A/en
Application granted granted Critical
Publication of CN114723590B publication Critical patent/CN114723590B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The invention discloses a knowledge tracking method, a knowledge tracking system, a knowledge tracking device and a knowledge tracking storage medium, and relates to the technical field of computers. The knowledge tracking method facing the group comprises the following steps: acquiring a first expression matrix of the student knowledge answer state, wherein the first expression matrix comprises a student group dimension, a knowledge point dimension and a time dimension; processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix; respectively carrying out spatial domain conversion processing on the plurality of first characteristic matrixes to obtain second expression matrixes, wherein the second expression matrixes comprise a plurality of second characteristic matrixes obtained after spatial domain conversion is carried out on the first characteristic matrixes; and modeling according to the second expression matrix to obtain a knowledge tracking prediction result. The method and the device can accurately predict the whole learning effect by taking student groups as objects, and carry out knowledge tracking.

Description

Group-oriented knowledge tracking method, system, device and storage medium
Technical Field
The invention relates to the technical field of computers, in particular to a group-oriented knowledge tracking method, a group-oriented knowledge tracking system, a group-oriented knowledge tracking device and a storage medium.
Background
Knowledge tracking is a basic and key task for supporting intelligent education service application, and aims to monitor the constantly-developed knowledge states of students, so that adaptive learning experience is provided for each student, the learning time is reasonably configured, and the purpose of improving the teaching quality and efficiency is achieved.
At present, a series of machine learning methods facing sequence modeling are adopted for knowledge tracking, and the aim of dynamically predicting the knowledge state of students by using learning interaction data is fulfilled. However, the knowledge tracking method in the related art considers students as independent and unrelated individuals in the learner modeling part, and ignores the influence of the relationship between learners and the group on the individuals, so that the knowledge tracking accuracy is poor in the learner model.
Disclosure of Invention
The present invention is directed to solving at least one of the problems of the prior art. Therefore, the invention provides a group-oriented knowledge tracking method, a group-oriented knowledge tracking system, a group-oriented knowledge tracking device and a storage medium, which can improve the accuracy of student group knowledge tracking.
In one aspect, an embodiment of the present invention provides a group-oriented knowledge tracking method, including the following steps:
acquiring a first representation matrix of student knowledge answer states, wherein the first representation matrix comprises a student group dimension, a knowledge point dimension and a time dimension;
processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix;
respectively carrying out spatial domain conversion processing on the first feature matrixes to obtain second expression matrixes, wherein the second expression matrixes comprise a plurality of second feature matrixes obtained after spatial domain conversion is carried out on the first feature matrixes;
and modeling according to the second expression matrix to obtain a knowledge tracking prediction result.
According to some embodiments of the invention, the processing the first representation matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps includes:
dividing the first expression matrix in a time dimension by taking a time step as a unit to obtain a plurality of answer matrixes of students and knowledge points;
and processing the answer matrix by adopting a generalized matrix decomposition model to obtain the student group characteristic representation matrix and the knowledge point characteristic representation matrix.
According to some embodiments of the present invention, the obtaining the second representation matrix by separately performing spatial-domain conversion on the plurality of first feature matrices includes:
acquiring a first spatial domain conversion matrix and a second spatial domain conversion matrix;
obtaining a first conversion matrix according to the student group characteristic representation matrix and the first spatial domain conversion matrix;
obtaining a second conversion matrix according to the knowledge point feature representation matrix and the second spatial domain conversion matrix;
and obtaining the second feature matrix according to the first conversion matrix and the second conversion matrix.
According to some embodiments of the invention, the first conversion matrix is obtained by the following formula:
Figure BDA0003572669160000021
wherein ,TSRepresents a first spatial-domain conversion matrix,
Figure BDA0003572669160000022
n represents the dimensional parameter of the embedded vector, S represents the number of students in the population,
Figure BDA0003572669160000023
a student group characteristic representation matrix representing a time step t;
the second transformation matrix is obtained by the following formula:
Figure BDA0003572669160000024
wherein ,TKRepresents a first spatial-domain conversion matrix,
Figure BDA0003572669160000025
n denotes the dimensional parameter of the embedded vector, K denotes the number of knowledge points,
Figure BDA0003572669160000026
a knowledge point feature representation matrix representing the time step t;
the second feature matrix is obtained by the following formula:
Figure BDA0003572669160000027
wherein ,
Figure BDA0003572669160000028
and a second feature matrix representing the time step t, wherein the second feature matrix is used for representing the interaction state of the student group and the knowledge point.
According to some embodiments of the invention, the modeling according to the second representation matrix to obtain the knowledge tracking prediction result comprises the following steps:
respectively carrying out knowledge point mastering level prediction processing according to the second feature matrixes to obtain a plurality of prediction vectors;
constructing an input sequence from a plurality of said prediction vectors;
and inputting the input sequence into a time cycle neural network for modeling to obtain a knowledge tracking prediction result.
According to some embodiments of the present invention, the obtaining a plurality of prediction vectors by performing the knowledge point mastering level prediction processing according to the plurality of second feature matrices respectively comprises:
inputting the second feature matrix corresponding to the time step t into a double-layer neural network to obtain a knowledge point predicted value corresponding to the time step t;
carrying out binarization processing on the knowledge point predicted value to obtain the predicted vector;
wherein the two-layer neural network is represented as:
Figure BDA0003572669160000029
wherein sigmoid represents an output layer of the double-layer neural network, W is a weight matrix, and b is a bias vector.
According to some embodiments of the invention, the time-cycled neural network is a long-short term memory network;
the input sequence is characterized as seq (a) ═ a1,A2,...,A|T|}, wherein
Figure BDA0003572669160000031
T represents the time dimension, AtRepresents the prediction vector corresponding to the time step t,
Figure BDA0003572669160000032
representing the mastery condition of the kth knowledge point in the prediction vector, wherein K represents the number of the knowledge points;
the output of the long-term and short-term memory network is a knowledge tracking prediction result P|T+1|，
Figure BDA0003572669160000033
wherein ,
Figure BDA0003572669160000034
and the predicted value of the mastery level of the student group on the ith knowledge point at the time step of | T | +1 is represented.
On the other hand, the embodiment of the invention also provides a knowledge tracking system facing to the group, which comprises:
the system comprises a first module, a second module and a third module, wherein the first module is used for acquiring a first representation matrix of student knowledge answer states, and the first representation matrix comprises a student group dimension, a knowledge point dimension and a time dimension;
the second module is used for processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix;
a third module, configured to perform spatial domain conversion on the multiple first feature matrices respectively to obtain second representation matrices, where the second representation matrices include multiple second feature matrices obtained by performing spatial domain conversion on the first feature matrices;
and the fourth module is used for modeling according to the second expression matrix to obtain a knowledge tracking prediction result.
On the other hand, an embodiment of the present invention further provides a group-oriented knowledge tracking apparatus, including:
at least one processor;
at least one memory for storing at least one program;
when executed by the at least one processor, cause the at least one processor to implement the population-oriented knowledge tracking method as described above.
In another aspect, the embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions for causing a computer to execute the group-oriented knowledge tracking method as described above.
The technical scheme of the invention at least has one of the following advantages or beneficial effects: the method comprises the steps of obtaining a first expression matrix including student group dimensions, knowledge point dimensions and time dimensions and representing student knowledge answer states, processing the first expression matrix according to the time dimensions to obtain a first feature matrix of a plurality of time steps, processing the obtained feature matrix including the student group feature expression matrix and the knowledge point feature expression matrix, respectively performing spatial domain conversion on the first feature matrices to obtain a second expression matrix including a plurality of second feature matrices, enabling the second feature matrix to represent interaction states of student group features and knowledge point features, modeling according to the second expression matrix to obtain a knowledge tracking prediction result, comprehensively considering influences among student groups in the obtained knowledge tracking prediction result, and achieving high accuracy of knowledge tracking of the student groups.
Drawings
FIG. 1 is a flow chart of a method for population-oriented knowledge tracking according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a segmentation and answer matrix decomposition of a first representation matrix based on a time dimension provided by an embodiment of the present invention;
FIG. 3 is a schematic diagram of space-domain mapping of a first representation matrix and a second representation matrix according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a process of processing a first feature matrix into a prediction vector according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating a long term memory network according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of a population-oriented knowledge tracking system provided by an embodiment of the present invention;
FIG. 7 is a schematic diagram of a population-oriented knowledge tracking device according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or components having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
In the description of the present invention, it should be understood that the orientation or positional relationship referred to in the description of the orientation, such as the upper, lower, left, right, etc., is based on the orientation or positional relationship shown in the drawings, and is only for convenience of description and simplicity of description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention.
In the description of the present invention, if there are first, second, etc. described, they are only used for distinguishing technical features, but they are not interpreted as indicating or implying relative importance or implicitly indicating the number of indicated technical features or implicitly indicating the precedence of the indicated technical features.
Studies have shown that in population learning, the relationships that exist from population to population, and from individual to individual, have an impact on learning motivation and efficacy. However, the current individual knowledge tracking method considers the learner modeling part as independent and unrelated individuals, neglects the relationship between learners and the influence of the group on the individuals, and thus the simplified learner model tends to lose the important information of the learner in the group learning environment.
In addition, it is not reasonable to simply superimpose the prediction results directly using the individual knowledge tracking as the prediction of the group knowledge state, because this scheme assumes each individual knowledge state as the prior condition of the group knowledge state. For example, an accuracy of α (0) is used<α<1) The individual knowledge tracking model predicts the knowledge state of class C consisting of N students, and the prediction accuracy beta is expressed as
Figure BDA0003572669160000041
wherein αiThe accuracy of the individual knowledge tracking model for predicting the knowledge state of the ith student in the class is assumed to be 10 even if alpha isiAt 0.9 (actually, the precision of the current knowledge tracking model is difficult to reach the level), the accuracy of the group prediction can only reach 0.349, and the accuracy of the prediction of the knowledge states of the students in the group is continuously reduced as the size of the group is increased.
In the practical teaching activities, most of offline teaching adopts the group teaching of class system, and based on the analysis, the group knowledge state is used as a reference standard to accurately measure the knowledge state of an individual learner in the personalized teaching.
Therefore, the embodiment of the invention provides a group-oriented knowledge tracking method to improve the accuracy of student group knowledge tracking. Referring to fig. 1, the knowledge tracking method facing the community according to the embodiment of the present invention includes, but is not limited to, step S110, step S120, step S130 and step S140.
Step S110, acquiring a first representation matrix of the student knowledge answer state, wherein the first representation matrix comprises student group dimensions, knowledge point dimensions and time dimensions;
step S120, processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix;
step S130, respectively carrying out spatial domain conversion processing on the plurality of first characteristic matrixes to obtain second expression matrixes, wherein the second expression matrixes comprise a plurality of second characteristic matrixes obtained after the spatial domain conversion processing is carried out on the first characteristic matrixes;
and step S140, modeling is carried out according to the second expression matrix to obtain a knowledge tracking prediction result.
In the embodiment, a first expression matrix which comprises student group dimensions, knowledge point dimensions and time dimensions and represents student knowledge answer states is obtained, the first expression matrix is processed according to the time dimensions to obtain a first feature matrix with a plurality of time steps, the processed feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix, then a second expression matrix comprising a plurality of second feature matrices is obtained by respectively carrying out spatial domain conversion on the first feature matrices, so that the second feature matrix can represent interaction states of student group features and knowledge point features, a knowledge tracking prediction result is obtained by modeling according to the second expression matrix, the influence among student groups is comprehensively considered in the obtained knowledge tracking prediction result, and the accuracy of knowledge tracking of the student groups is high.
According to some embodiments of the invention, step S120 further includes, but is not limited to, the following steps:
step S210, with time step as a unit, carrying out time dimension segmentation on the first expression matrix to obtain a plurality of answer matrixes of students and knowledge points;
and step S220, processing the answer matrix by adopting a generalized matrix decomposition model to obtain a student group characteristic representation matrix and a knowledge point characteristic representation matrix.
In the present embodiment, the knowledge response states of the student population can be observed and recorded through three dimensions, namely, a student population dimension (S dimension), a knowledge point dimension (K dimension), and a time dimension (T dimension), so as to obtain the first representation matrix in the individual spatial domain as shown in fig. 2. As shown in fig. 2, the first representation matrix is cut by taking time as a unit to obtain an answer matrix of students and knowledge points at each historical time step t
Figure BDA0003572669160000051
If the number of students in the group is | S |, and the number of knowledge points is | K |, then
Figure BDA0003572669160000052
Dimension (d) of
Figure BDA0003572669160000053
Figure BDA0003572669160000054
The row of (a) may represent the answer situation or the answer score of each student in the group to each knowledge point at the current time step t, and the column of the matrix may represent the situation where each knowledge point is answered by each student in the group. Illustratively, the jth row of the answer matrix represents student sjIth list indicating known points kiWhen the data value at the intersection of the jth row and ith column is 1, it indicates that the student s is a studentjCorrectly answer the knowledge point kiAnd if 0, it means student sjKnowledge point k without correct answeri。
In some of the other embodiments, the first and second electrodes are,
Figure BDA0003572669160000061
the column (b) may also represent the answer condition or the answer score of each student in the group to each knowledge point at the current time step t, and the industry of the matrix may represent that each knowledge point is grouped by each studentThe condition of the answer.
Answer matrix
Figure BDA0003572669160000062
Obtaining a student group characteristic representation Matrix under an individual space domain at t time step through the processing of a Generalized Matrix Factorization model (Generalized Matrix Factorization)
Figure BDA0003572669160000063
And knowledge point feature representation matrix
Figure BDA0003572669160000064
wherein
Figure BDA0003572669160000065
n is the dimension parameter of the embedding vector.
According to some embodiments of the present invention, step S130 further includes, but is not limited to, the following steps:
step S310, a first spatial domain conversion matrix and a second spatial domain conversion matrix are obtained;
step S320, obtaining a first conversion matrix according to the student group characteristic representation matrix and the first spatial domain conversion matrix;
step S330, a second conversion matrix is obtained according to the knowledge point characteristic representation matrix and the second spatial domain conversion matrix;
step S340, a second feature matrix is obtained according to the first transformation matrix and the second transformation matrix.
In this embodiment, referring to fig. 3, it is necessary to map the first representation matrix characterizing the individual spatial domain to the second representation matrix characterizing the group spatial domain. Specifically, referring to fig. 4, a first transformation matrix is obtained according to the student population characteristic representation matrix and the first spatial domain transformation matrix, and the first transformation matrix is obtained by formula (1):
Figure BDA0003572669160000066
wherein ,TSRepresenting a first spatial domainThe matrix is converted into a matrix of values,
Figure BDA0003572669160000067
n represents the dimensional parameter of the embedded vector, S represents the number of students in the population,
Figure BDA0003572669160000068
a student group characteristic representation matrix representing the time step t;
obtaining a second transformation matrix according to the knowledge point feature representation matrix and the second spatial domain transformation matrix, wherein the second transformation matrix is obtained through a formula (2):
Figure BDA0003572669160000069
wherein ,TKRepresents a first spatial-domain conversion matrix,
Figure BDA00035726691600000610
n represents the dimension parameter of the embedded vector, the feature of the knowledge points is ensured to represent the dimension of the matrix unchanged in the operation process, the subscript K represents the number of the knowledge points,
Figure BDA00035726691600000611
a knowledge point feature representation matrix representing the time step t;
obtaining a second feature matrix according to the first conversion matrix and the second conversion matrix, wherein the second feature matrix is obtained by a formula (3):
Figure BDA00035726691600000612
wherein ,
Figure BDA00035726691600000613
a second feature matrix representing the time step t,
Figure BDA00035726691600000614
the second feature matrix is used for representing the student population under the population space domainThe interaction state of the knowledge points.
In some further embodiments, the first spatial-domain conversion matrix and the second spatial-domain conversion matrix may each be trained by minimizing an error between real values and an output of the conversion model
According to some embodiments of the invention, step S140 further includes, but is not limited to, step S410, step S420, and step S430.
Step S410, knowledge point mastering level prediction processing is carried out according to the second feature matrixes respectively to obtain a plurality of prediction vectors;
step S420, an input sequence is constructed according to a plurality of prediction vectors;
and step S430, inputting the input sequence into a time-cycle neural network for modeling to obtain a knowledge tracking prediction result.
Further, step S410 further includes, but is not limited to, step S510 and step S520.
Step S510, inputting a second feature matrix corresponding to the time step t into a double-layer neural network to obtain a knowledge point predicted value corresponding to the time step t;
the two-layer neural network is represented by equation (4):
Figure BDA0003572669160000071
wherein sigmoid represents an output layer of the double-layer neural network, W is a weight matrix, and b is a bias vector.
The double-layer neural network can be optimized by taking the binary cross entropy as a loss function, wherein the binary cross entropy is taken as the loss function as shown in formula (5):
Figure BDA0003572669160000072
wherein ,riThe number of students who answer the knowledge point i is represented as the proportion of the number of the groups.
And step S520, carrying out binarization processing on the knowledge point predicted value to obtain a predicted vector.
The threshold value theta is set, and the value range of theta can be set between 0.5 and 1. With continued reference to FIG. 4, if
Figure BDA0003572669160000073
Indicates that the student group masters the knowledge point i to obtain aiIf 1, then
Figure BDA0003572669160000074
Indicating that the student group does not master the knowledge point i, namely ai＝0。
After each historical time step is processed, the prediction vector A of the mastery condition of each knowledge point of the student group at each time step t is obtainedt，
Figure BDA0003572669160000075
Sorting the plurality of prediction vectors in time steps to obtain a time-sorted input sequence seq (a) { a }1,A2,...,A|T|}。
Further, the time-cycled neural network in step S430 may employ a long-short term memory network (LSTM), which is shown in fig. 5. Modeling student population knowledge states using LSTM, the input sequence is seq (A) ═ { A1,A2,...,A|T|}, wherein
Figure BDA0003572669160000076
T represents the time dimension, AtRepresents the prediction vector corresponding to the time step t,
Figure BDA0003572669160000077
the knowledge point K in the prediction vector is shown, and K represents the number of knowledge points.
htThe hidden state of the student group for mastering all knowledge points at the time step t, h0Is the initialization input for the LSTM.
After completing the modeling training of the input sequence with length of | T |, the output of the long-short term memory network is the prediction of knowledge trackingResults P|T+1|，
Figure BDA0003572669160000078
wherein ,
Figure BDA0003572669160000079
and the grasping level prediction value of the student group to the ith knowledge point at the time step of | T | +1 is represented.
On the other hand, an embodiment of the present invention further provides a group-oriented knowledge tracking system, referring to fig. 6, including:
the system comprises a first module, a second module and a third module, wherein the first module is used for acquiring a first representation matrix of the student knowledge answer state, and the first representation matrix comprises a student group dimension, a knowledge point dimension and a time dimension;
the second module is used for processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix;
a third module, configured to perform spatial domain conversion processing on the multiple first feature matrices respectively to obtain second representation matrices, where the second representation matrices include multiple second feature matrices obtained by performing spatial domain conversion on the first feature matrices;
and the fourth module is used for modeling according to the second expression matrix to obtain a knowledge tracking prediction result.
It is to be understood that the contents of the embodiment of the knowledge tracking method for group oriented are all applicable to the embodiment of the present system, the functions implemented by the embodiment of the present system are the same as the embodiment of the knowledge tracking method for group oriented, and the advantageous effects achieved by the embodiment of the knowledge tracking method for group oriented are also the same as the advantageous effects achieved by the embodiment of the knowledge tracking method for group oriented.
Referring to fig. 7, fig. 7 is a schematic diagram of a knowledge tracking apparatus facing a group according to an embodiment of the present invention. The population-oriented knowledge tracking device of the embodiment of the invention comprises one or more control processors and a memory, and one control processor and one memory are taken as an example in fig. 7.
The control processor and the memory may be connected by a bus or other means, as exemplified by the bus connection in fig. 7.
The memory, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer executable programs. Further, the memory may include high speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory remotely located from the control processor, and the remote memory may be connected to the population-oriented knowledge tracking device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
Those skilled in the art will appreciate that the device configuration shown in FIG. 7 does not constitute a limitation of a population-oriented knowledge tracking device, and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.
The non-transitory software programs and instructions required to implement the swarm-oriented knowledge tracking method applied to the swarm-oriented knowledge tracking device in the above embodiments are stored in a memory and, when executed by a control processor, perform the swarm-oriented knowledge tracking method applied to the swarm-oriented knowledge tracking device in the above embodiments.
Furthermore, an embodiment of the present invention also provides a computer-readable storage medium storing computer-executable instructions, which are executed by one or more control processors, and can cause the one or more control processors to execute the group-oriented knowledge tracking method in the above method embodiment.
One of ordinary skill in the art will appreciate that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.
The embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention.

Claims (10)

1. A population-oriented knowledge tracking method, comprising the steps of:
acquiring a first representation matrix of student knowledge answer states, wherein the first representation matrix comprises a student group dimension, a knowledge point dimension and a time dimension;
processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix;
respectively carrying out spatial domain conversion processing on the first feature matrixes to obtain second expression matrixes, wherein the second expression matrixes comprise a plurality of second feature matrixes obtained after spatial domain conversion is carried out on the first feature matrixes;
and modeling according to the second expression matrix to obtain a knowledge tracking prediction result.
2. The population-oriented knowledge tracking method of claim 1, wherein the processing the first representation matrix according to the time dimension to obtain a first feature matrix for a plurality of time steps comprises:
dividing the first expression matrix in a time dimension by taking a time step as a unit to obtain a plurality of answer matrixes of students and knowledge points;
and processing the answer matrix by adopting a generalized matrix decomposition model to obtain the student group characteristic representation matrix and the knowledge point characteristic representation matrix.
3. The population-oriented knowledge tracking method according to claim 2, wherein the step of performing spatial-domain transformation on the plurality of first feature matrices to obtain second representation matrices comprises:
acquiring a first spatial domain conversion matrix and a second spatial domain conversion matrix;
obtaining a first conversion matrix according to the student group characteristic representation matrix and the first spatial domain conversion matrix;
obtaining a second conversion matrix according to the knowledge point feature representation matrix and the second spatial domain conversion matrix;
and obtaining the second feature matrix according to the first conversion matrix and the second conversion matrix.
4. The population-oriented knowledge tracking method of claim 3, wherein the first transformation matrix is obtained by the formula:
Figure FDA0003572669150000011
wherein ,TSRepresents a first spatial-domain conversion matrix,
Figure FDA0003572669150000012
n represents the dimensional parameter of the embedded vector, S represents the number of students in the population,
Figure FDA0003572669150000013
a student group characteristic representation matrix representing the time step t;
the second transformation matrix is obtained by the following formula:
Figure FDA0003572669150000014
wherein ,TKRepresents a first spatial-domain conversion matrix,
Figure FDA0003572669150000015
n denotes the dimensional parameter of the embedded vector, K denotes the number of knowledge points,
Figure FDA0003572669150000016
a knowledge point feature representation matrix representing the time step t;
the second feature matrix is obtained by the following formula:
Figure FDA0003572669150000021
wherein ,
Figure FDA0003572669150000022
and a second feature matrix representing the time step t, wherein the second feature matrix is used for representing the interaction state of the student group and the knowledge point.
5. The population-oriented knowledge tracking method of claim 4, wherein the modeling from the second representation matrix to obtain knowledge tracking predictions comprises:
respectively carrying out knowledge point mastering level prediction processing according to the second feature matrixes to obtain a plurality of prediction vectors;
constructing an input sequence from a plurality of said prediction vectors;
and inputting the input sequence into a time-cycle neural network for modeling and obtaining a knowledge tracking prediction result.
6. The population-oriented knowledge tracking method according to claim 5, wherein the step of performing knowledge point grasp level prediction processing on the basis of the second feature matrices to obtain prediction vectors comprises the steps of:
inputting the second characteristic matrix corresponding to the time step t into a double-layer neural network to obtain a knowledge point predicted value corresponding to the time step t;
carrying out binarization processing on the knowledge point predicted value to obtain the predicted vector;
wherein the two-layer neural network is represented as:
Figure FDA0003572669150000023
wherein sigmoid represents an output layer of the double-layer neural network, W is a weight matrix, and b is a bias vector.
7. The population-oriented knowledge tracking method of claim 6, wherein the time-cycled neural network is a long-short term memory network;
the input sequence is characterized as seq (a) ═ a1,A2,...，A|T|}, wherein
Figure FDA0003572669150000024
When T representsMiddle dimension, AtRepresents the prediction vector corresponding to the time step t,
Figure FDA0003572669150000025
representing the mastery condition of the kth knowledge point in the prediction vector, wherein K represents the number of the knowledge points;
the output of the long-term and short-term memory network is a knowledge tracking prediction result P|T+1|，
Figure FDA0003572669150000026
wherein ,
Figure FDA0003572669150000027
and the predicted value of the mastery level of the student group on the ith knowledge point at the time step of | T | +1 is represented.
8. A population-oriented knowledge tracking system, comprising:
the system comprises a first module, a second module and a third module, wherein the first module is used for acquiring a first representation matrix of student knowledge answer states, and the first representation matrix comprises a student group dimension, a knowledge point dimension and a time dimension;
the second module is used for processing the first expression matrix according to the time dimension to obtain a first feature matrix of a plurality of time steps, wherein the feature matrix comprises a student group feature expression matrix and a knowledge point feature expression matrix;
a third module, configured to perform spatial domain conversion processing on the multiple first feature matrices respectively to obtain second representation matrices, where the second representation matrices include multiple second feature matrices obtained by performing spatial domain conversion on the first feature matrices;
and the fourth module is used for modeling according to the second expression matrix to obtain a knowledge tracking prediction result.
9. A population-oriented knowledge tracking device, comprising:
at least one processor;
at least one memory for storing at least one program;
when executed by the at least one processor, cause the at least one processor to implement the population-oriented knowledge tracking method of any one of claims 1 to 7.
10. A computer readable storage medium having stored therein a processor executable program for implementing the population-oriented knowledge tracking method of any one of claims 1 to 7 when executed by the processor.
CN202210323642.XA 2022-03-30 2022-03-30 Group-oriented knowledge tracking method, system, device and storage medium Active CN114723590B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210323642.XA CN114723590B (en) 2022-03-30 2022-03-30 Group-oriented knowledge tracking method, system, device and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210323642.XA CN114723590B (en) 2022-03-30 2022-03-30 Group-oriented knowledge tracking method, system, device and storage medium

Publications (2)

Publication Number Publication Date
CN114723590A true CN114723590A (en) 2022-07-08
CN114723590B CN114723590B (en) 2023-04-28

Family

ID=82239993

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210323642.XA Active CN114723590B (en) 2022-03-30 2022-03-30 Group-oriented knowledge tracking method, system, device and storage medium

Country Status (1)

Country Link
CN (1) CN114723590B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108052672A (en) * 2017-12-29 2018-05-18 北京师范大学 Promote structural knowledge map construction system and method using group study behavior
CN109840595A (en) * 2019-02-26 2019-06-04 武汉大学 A kind of knowledge method for tracing based on group study behavior feature
CN110428010A (en) * 2019-08-05 2019-11-08 中国科学技术大学 Knowledge method for tracing
CN110807469A (en) * 2019-09-19 2020-02-18 华中师范大学 Knowledge tracking method and system integrating long-time memory and short-time memory with Bayesian network
CN113077184A (en) * 2021-04-27 2021-07-06 上海松鼠课堂人工智能科技有限公司 Data adjustment method, device, equipment and medium based on Bayesian knowledge tracking
CN113255433A (en) * 2021-04-06 2021-08-13 北京迈格威科技有限公司 Model training method, device and computer storage medium
CN113344053A (en) * 2021-05-29 2021-09-03 华中师范大学 Knowledge tracking method based on examination question heteromorphic image representation and learner embedding
CN113361685A (en) * 2021-05-25 2021-09-07 华中师范大学 Knowledge tracking method and system based on learner knowledge state evolution expression
US20210374419A1 (en) * 2020-06-02 2021-12-02 Stats Llc Semi-Supervised Action-Actor Detection from Tracking Data in Sport
CN113947262A (en) * 2021-11-25 2022-01-18 陕西师范大学 Knowledge tracking method based on different composition learning fusion learning participation state

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108052672A (en) * 2017-12-29 2018-05-18 北京师范大学 Promote structural knowledge map construction system and method using group study behavior
CN109840595A (en) * 2019-02-26 2019-06-04 武汉大学 A kind of knowledge method for tracing based on group study behavior feature
CN110428010A (en) * 2019-08-05 2019-11-08 中国科学技术大学 Knowledge method for tracing
CN110807469A (en) * 2019-09-19 2020-02-18 华中师范大学 Knowledge tracking method and system integrating long-time memory and short-time memory with Bayesian network
US20210374419A1 (en) * 2020-06-02 2021-12-02 Stats Llc Semi-Supervised Action-Actor Detection from Tracking Data in Sport
CN113255433A (en) * 2021-04-06 2021-08-13 北京迈格威科技有限公司 Model training method, device and computer storage medium
CN113077184A (en) * 2021-04-27 2021-07-06 上海松鼠课堂人工智能科技有限公司 Data adjustment method, device, equipment and medium based on Bayesian knowledge tracking
CN113361685A (en) * 2021-05-25 2021-09-07 华中师范大学 Knowledge tracking method and system based on learner knowledge state evolution expression
CN113344053A (en) * 2021-05-29 2021-09-03 华中师范大学 Knowledge tracking method based on examination question heteromorphic image representation and learner embedding
CN113947262A (en) * 2021-11-25 2022-01-18 陕西师范大学 Knowledge tracking method based on different composition learning fusion learning participation state

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
ZHENGYANG WU 等: ""Exercise recommendation based on knowledge concept prediction"", 《ELSEVIER》 *
李浩君 等: ""基于自注意力机制和双向GRU神经网络的深度知识追踪优化模型"", 《计算机应用研究》 *
闾汉原等: "基于"态度"的知识追踪模型及集成技术", 《徐州师范大学学报(自然科学版)》 *

Also Published As

Publication number Publication date
CN114723590B (en) 2023-04-28

Similar Documents

Publication Publication Date Title
CN110807469B (en) Knowledge tracking method and system integrating long-time memory and short-time memory with Bayesian network
US20170103308A1 (en) Acceleration of convolutional neural network training using stochastic perforation
CN109919183B (en) Image identification method, device and equipment based on small samples and storage medium
CN110135562B (en) Distillation learning method, system and device based on characteristic space change
CN112116092A (en) Interpretable knowledge level tracking method, system and storage medium
CN113361685B (en) Knowledge tracking method and system based on learner knowledge state evolution expression
CN111159419A (en) Knowledge tracking data processing method, system and storage medium based on graph convolution
Xiong et al. Predicting learning status in MOOCs using LSTM
CN112860847B (en) Video question-answer interaction method and system
CN115329959A (en) Learning target recommendation method based on double-flow knowledge embedded network
CN110704668B (en) Grid-based collaborative attention VQA method and device
CN115545160A (en) Knowledge tracking method and system based on multi-learning behavior cooperation
CN114974397A (en) Training method of protein structure prediction model and protein structure prediction method
CN113408621B (en) Rapid simulation learning method, system and equipment for robot skill learning
KR20180061998A (en) System and method for diagnosing mass attributes of leaner
CN112818100B (en) Knowledge tracking method and system for integrating question difficulty
CN113705402A (en) Video behavior prediction method, system, electronic device and storage medium
CN114723590B (en) Group-oriented knowledge tracking method, system, device and storage medium
CN114819325A (en) Score prediction method, system, device and storage medium based on graph neural network
CN114971066A (en) Knowledge tracking method and system integrating forgetting factor and learning ability
CN116578694B (en) Disentangled knowledge tracking method, system, device and medium
KR20210105272A (en) Pre-training modeling system and method for predicting educational factors
KR102635769B1 (en) Learning effect estimation device, learning effect estimation method, program
CN115700608A (en) Data processing method and device, computer equipment and storage medium
JPH0844699A (en) Nonlinear time-series data predicting device

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
GR01 Patent grant
GR01 Patent grant