Student ability profiling method, apparatus, electronic device, and storage medium

Abstract

The application discloses a student ability portrait method and device, electronic equipment and a storage medium, and belongs to the technical field of computers, wherein the student ability portrait method comprises the following steps: modeling a historical answering process of a student to obtain a first historical state representation of the student; obtaining a second historical state representation of the student based on historical answering records of the student; fusing the first historical state representation and the second historical state representation of the student to obtain a fused historical state representation; performing multi-dimensional ability mapping on the fused historical state representation to obtain an ability representation of the student; and obtaining a portrait representation of the student based on the ability representation and a knowledge point representation of the student. The application can more accurately fit the student ability, predict the student performance, provide a basis for adaptive learning and personalized learning according to individual differences, and greatly improve the student learning efficiency.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method, apparatus, electronic device, and storage medium for profiling student abilities. Background Technology

[0002] With the development of information technology, the informatization and intelligentization of education have gradually gained importance, especially in how to provide personalized educational resources for students based on educational big data, which has become a crucial need for schools, parents, and students. Accurately characterizing student status is key to providing personalized learning resources, achieving individualized instruction, and promoting both reduced workload and increased efficiency. Therefore, the automatic characterization and description of student abilities based on historical records has become a fundamental and important research question. Existing technologies use cognitive diagnostic models to model student abilities, which are based on item response theory or Bayesian theory. These models typically have a small number of parameters, poor data utilization, and difficulty in extracting information from large amounts of educational data, resulting in poor accuracy in complex scenarios. Furthermore, existing student modeling models usually model student abilities as a scalar, a representation method with limited modeling capabilities, making it difficult to accurately characterize students' knowledge and abilities. Summary of the Invention

[0003] This invention provides a student ability profiling method, device, electronic device, and storage medium to solve the problems of poor data utilization, difficulty in mining information from large amounts of educational data, poor accuracy in complex scenarios, and limited modeling capabilities that make it difficult to accurately depict students' knowledge and abilities in the existing technology. It realizes multi-dimensional mapping of student abilities, more accurately fits student abilities, and predicts student performance.

[0004] In a first aspect, the present invention provides a method for profiling student abilities, comprising:

[0005] Model the student's historical answering process to obtain the student's first historical state representation;

[0006] Based on the student's historical answer records, obtain the student's second historical state representation;

[0007] By fusing the first and second historical state representations of the student, a fused historical state representation is obtained.

[0008] A multidimensional capability mapping is performed on the fused historical state representation to obtain the student's capability representation;

[0009] Based on the student's ability representation and knowledge point representation, a profile representation of the student is obtained.

[0010] In some embodiments, modeling the student's historical answering process to obtain a first historical state representation of the student includes:

[0011] Based on recurrent neural networks, the student's historical answering process is modeled to obtain the student's first historical state representation.

[0012] In some embodiments, obtaining the student's second historical state representation based on the student's historical answer records includes:

[0013] Based on an inductive algorithm, the student's answers to each knowledge point in the historical answer record are statistically analyzed, and the student's historical scores are accumulated in an exponential decay manner to obtain the student's second historical state representation.

[0014] In some embodiments, performing multidimensional capability mapping on the fused historical state representation to obtain the student's capability representation includes:

[0015] Determine the student knowledge matrix;

[0016] Based on the student knowledge matrix, a multidimensional mapping is performed on the fused historical state representation to obtain the student's ability representation.

[0017] In some embodiments, obtaining a student profile representation based on the student's ability representation and knowledge point representation includes:

[0018] The student's ability representation and knowledge point representation are fused together, and an initial profile representation of the student in terms of knowledge points is obtained through an ability mapping network.

[0019] In some embodiments, the method further includes:

[0020] By combining the error factors and guessing factors in the student's learning process, the initial profile representation is processed to obtain the final profile representation of the student.

[0021] In some embodiments, the process of determining the capability mapping network includes:

[0022] Obtain student ability representation samples and profile representation samples;

[0023] The capability representation samples are input into the initial capability mapping network to obtain the predicted profile representation;

[0024] The initial capability mapping network is obtained by iterating the parameters of the predicted profile representation and the profile representation samples.

[0025] Secondly, the present invention provides a student ability profiling device, comprising:

[0026] The first acquisition unit models the student's historical answering process to obtain the student's first historical state representation;

[0027] The second acquisition unit is used to acquire the student's second historical state representation based on the student's historical answer records;

[0028] The fusion unit is used to fuse the student's first historical state representation and second historical state representation to obtain a fused historical state representation.

[0029] A capability mapping unit is used to perform multi-dimensional capability mapping on the fused historical state representation to obtain the student's capability representation.

[0030] A profile unit is used to obtain a profile representation of the student based on the student's ability representation and knowledge point representation.

[0031] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the student competency profiling method as described in any of the first aspects.

[0032] Fourthly, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the student competency profiling method as described in any of the first aspects.

[0033] This invention provides a student ability profiling method, device, electronic device, and storage medium. It obtains a first historical state representation of a student by modeling their historical answering process; a second historical state representation is obtained based on the student's historical answering records; the first and second historical state representations are merged to obtain a fused historical state representation; multi-dimensional ability mapping is performed on the fused historical state representation to obtain the student's ability representation; and a student profile representation is obtained based on the student's ability representation and knowledge point representation. This enables a more accurate fit to student abilities, predicts student performance, and provides a foundation for adaptive and personalized learning tailored to individual needs, significantly improving student learning efficiency. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0035] Figure 1This is an application environment diagram illustrating the operational capability of the student competency profiling method provided in one embodiment of the present invention.

[0036] Figure 2 This is a flowchart illustrating a student ability profiling method according to an embodiment of the present invention;

[0037] Figure 3 This is a schematic diagram of the structure of a student ability profiling device provided in one embodiment of the present invention;

[0038] Figure 4 This is a schematic diagram of the physical structure of an electronic device provided in one embodiment of the present invention. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0040] The terms "first," "second," etc., used in the specification and claims of this invention are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of the invention can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, the first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0041] To address the shortcomings of existing technologies, such as poor data utilization, difficulty in extracting information from large amounts of educational data, poor accuracy in complex scenarios, and limited modeling capabilities that make it difficult to accurately characterize students' knowledge and abilities, this invention provides a student ability profiling method, device, electronic device, and storage medium. The following describes... Figures 1-4 The present invention will now be described.

[0042] The student ability profiling method provided by this invention can be applied to, for example... Figure 1 The application environment shown. Figure 1 This diagram illustrates the application environment in which the student ability profiling method provided in one embodiment of the present invention can operate. For example... Figure 1As shown, the application environment includes terminal 110 and server 120. Terminal 110 and server 120 communicate via a network, which can be a wireless communication network or a wired communication network. The number of terminals and servers is unlimited. The wireless communication network can include, but is not limited to, at least one of the following: WIFI (Wireless Fidelity) and Bluetooth. The wired communication network can include, but is not limited to, at least one of the following: wide area network, metropolitan area network, and local area network.

[0043] Terminal 110 includes various handheld devices, in-vehicle devices, wearable devices, computing devices, or other processing devices connected to a wireless modem with wireless or wired communication capabilities, such as mobile phones, tablets, desktop laptops, and smart devices capable of running applications, including the central console of a smart car, smartphones, etc.

[0044] Server 120 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0045] It should be noted that the student ability profiling method of the present invention can be implemented directly on the terminal 110, or directly on the server 120, or it can be completed on the server 120 and then sent to the terminal 110 by the server 120.

[0046] Terminal 110 or server 120 models the student's historical answer process to obtain the student's first historical state representation; based on the student's historical answer records, obtain the student's second historical state representation; merge the student's first and second historical state representations to obtain a merged historical state representation; perform multi-dimensional ability mapping on the merged historical state representation to obtain the student's ability representation; based on the student's ability representation and knowledge point representation, obtain the student's profile representation, thereby achieving a more accurate fit to the student's ability, predicting student performance, providing a foundation for adaptive learning and personalized learning that caters to individual needs, and greatly improving student learning efficiency.

[0047] Figure 2 This is a flowchart illustrating a student ability profiling method according to an embodiment of the present invention. Figure 2 As shown, a student ability profiling method is provided, including the following steps: step 200, step 201, step 202, step 203, and step 204. This method's steps are merely one possible implementation of the present invention.

[0048] Step 200: Model the student's historical answering process to obtain the student's first historical state representation;

[0049] Optionally, by simulating the memory and forgetting factors of the students during the history question-answering process, the changes in the students' knowledge state over time can be fitted, hidden information in the students' history question-answering process can be mined, and the students' first historical state representation H can be obtained. b The first historical state represents H. b This reflects the student's state during the history question-answering process.

[0050] Step 201: Based on the student's historical answer records, obtain the student's second historical state representation;

[0051] Optionally, the student's answers to each knowledge point in the historical answer record are statistically analyzed. By judging the student's mastery of each knowledge point within a preset time period, the student's answer score trend is analyzed to obtain the student's second historical state representation H. a The preset time period can be weekly, monthly, or semester-based. The student's second historical state is represented by H. a It can intuitively reflect the students' scoring trends on various knowledge points.

[0052] Step 202: Merge the first and second historical state representations of the student to obtain the merged historical state representation;

[0053] Optionally, the student's first historical state is represented as H. b Second historical state representation H a The fusion process is performed to obtain the fused historical state representation H.

[0054] It is understandable that H = H a +H b The student's final ability state consists of two parts, namely the student's first historical state representation H. b Second historical state representation H a The integrated historical state, represented by H, reflects the students' mastery of each knowledge point.

[0055] Step 203: Perform multi-dimensional capability mapping on the fused historical state representation to obtain the student's capability representation;

[0056] Alternatively, each student is an independent individual with individual differences. Using only one dimension to measure each student's ability may lead to misjudgment and is not conducive to individualized teaching. Therefore, it is necessary to measure each student's ability from more dimensions.

[0057] Optionally, multidimensional ability mapping refers to using the student knowledge matrix to perform multidimensional mapping on the fused historical state representation, thereby enabling a deeper description of student abilities and uncovering more information representing student abilities.

[0058] The student knowledge matrix represents knowledge in multiple dimensions, such as the ability tested by the knowledge, the difficulty level of the knowledge, the subject attribute of the knowledge, and the grade attribute of the knowledge, which implies the ability requirements of the knowledge for students.

[0059] Let the student's first historical state be represented as H. b Second historical state representation H a The fusion process yields the fused historical state representation H, which is the student's ability representation P obtained through multi-dimensional ability mapping based on the student's mastery of each knowledge point.

[0060] Step 204: Based on the student's ability representation and knowledge point representation, obtain the student's profile representation.

[0061] Optionally, the student's ability representation P and knowledge point representation a are input into a multilayer neural network to predict the student profile and obtain the student's profile representation p. Compared with traditional cognitive diagnostic models, multilayer neural networks have better reasoning and modeling capabilities.

[0062] In this embodiment of the invention, a first historical state representation of the student is obtained by modeling the student's historical answering process; a second historical state representation of the student is obtained based on the student's historical answering records; the first and second historical state representations are fused, and the fused historical state representation is subjected to multi-dimensional ability mapping to obtain the student's ability representation; based on the student's ability representation and knowledge point representation, a student profile representation is obtained. This method can more accurately fit student abilities and predict student performance, providing a foundation for adaptive learning and personalized learning tailored to individual needs, and can greatly improve student learning efficiency.

[0063] It should be noted that each embodiment of the present invention can be freely combined, rearranged, or executed individually, and does not need to rely on or depend on a fixed execution order.

[0064] In some embodiments, modeling the student's historical answering process to obtain a first historical state representation of the student includes:

[0065] Based on recurrent neural networks, the student's historical answering process is modeled to obtain the student's first historical state representation.

[0066] Optionally, a Recurrent Neural Network (RNN) is a recurrent neural network that takes sequence data as input, recursively processes the sequence in the direction of its evolution, and all nodes are linked in a chain-like manner. Bidirectional RNNs (Bi-RNNs) and Long Short-Term Memory networks (LSTMs) are common types of RNNs. The main application of RNNs is processing and predicting sequence data.

[0067] Based on recurrent neural networks, specifically using Long Short-Term Memory (LSTM) networks, a common type of recurrent neural network, we model the student's historical answering process. The LSM network defines the hidden state h. t Hidden state h t This represents the student's knowledge state at a certain moment, as well as the cell state c. t To store historical information, a three-gate structure was constructed to update and calculate student status time-by-time. The specific process is as follows:

[0068] i t =σ(Z) xi x t +Z hi h t-1 +b i )

[0069] f t =σ(Z) xf x t +Z hf h t-1 +b f )

[0070] o t =σ(Z) xo x t +Z ho h t-1 +b o )

[0071] c t =f t ·c t-1 +i t ·tanh(Z xc x t +Z hc h t-1 +b c )

[0072] h t =o t ·tanh(c t )

[0073] In the above formula, x t This represents the sequence information input to the Long Short-Term Memory (LSTM) network module; in this case, it's the student's historical answer sequence. σ and tanh are two commonly used nonlinear activation functions. t f t o t These represent the input gate, forget gate, and output gate, respectively. The input gate controls the selective memorization of input information, while the forget gate controls the processing of the memorized information c passed from the previous node. t-1 Partial forgetting is performed, while the output gate determines how much information can be used as the current state h. t The output of the Long Short-Term Memory (LSTM) network module can simulate the memory and forgetting factors of students in the historical question-answering process, thereby fitting the change process of students' knowledge status over time. Through this LSM network, the user's question-answering sequence (s) is processed. t ,a t ,q t The modeling uses a triple to represent the user's answer at time t, which is attributed to knowledge point a. t The question q t The answer was s t Finally, the first historical state representation H of the student is obtained through this network. b .

[0074] In this embodiment of the invention, a recurrent neural network is used to model the student's historical answering process, obtain the student's first historical state representation, and obtain the hidden information of the student in the historical answering process, thereby more accurately representing the student's ability.

[0075] In some embodiments, obtaining the student's second historical state representation based on the student's historical answer records includes:

[0076] Based on an inductive algorithm, the student's answers to each knowledge point in the historical answer record are statistically analyzed, and the student's historical scores are accumulated in an exponential decay manner to obtain the student's second historical state representation.

[0077] Optionally, based on an inductive algorithm, the student's answers to each knowledge point in the historical answer record are statistically analyzed. Simultaneously, to preserve the student's temporal information, the student's historical scores are accumulated using an exponential decay method. The student's inductive historical state at time t for a certain knowledge point k can be represented as:

[0078] h k,t =h k,t-1 *α+score t

[0079] Where *score* represents the student's score at time *t* for knowledge point *k*. *α* is a coefficient that can be flexibly set according to requirements, for example, a coefficient of 0.2. Finally, based on an inductive algorithm, the student's second historical state representation *H* can be obtained. a .

[0080] In this embodiment of the invention, based on an inductive algorithm, the student's answers to each knowledge point in the historical answer record are statistically analyzed, and the student's historical scores are accumulated in an exponential decay manner to obtain the student's second historical state representation, which intuitively reflects the student's score trend for each knowledge point.

[0081] In some embodiments, performing multidimensional capability mapping on the fused historical state representation to obtain the student's capability representation includes:

[0082] Determine the student knowledge matrix;

[0083] Based on the student knowledge matrix, a multidimensional mapping is performed on the fused historical state representation to obtain the student's ability representation.

[0084] Optionally, the fused historical state representation is subjected to multidimensional mapping. This invention defines a student knowledge matrix M, which is a multidimensional matrix obtained through random initialization and training. Each row of this matrix represents knowledge in more dimensions, implying characteristics such as the abilities required of the student based on that knowledge.

[0085] Based on the student knowledge matrix M, a multidimensional mapping is performed on the fused historical state representation H to obtain the student's ability representation P, which is a multidimensional representation of the student's knowledge ability. For example, if the student knowledge matrix M is a 10*5 matrix and the fused historical state representation H is a 10*1 matrix, the student's ability representation P = H·M is obtained, where P is a 10*5 matrix. Compared to H, P has more dimensions and can describe the student's ability at a deeper level.

[0086] In this embodiment of the invention, a student knowledge matrix is ​​determined; based on the student knowledge matrix, the fused historical state representation is mapped in multiple dimensions to obtain the student's ability representation, thereby realizing the multi-dimensional mapping expansion of student abilities, which can describe student abilities more deeply and predict student performance.

[0087] In some embodiments, obtaining a student profile representation based on the student's ability representation and knowledge point representation includes:

[0088] The student's ability representation and knowledge point representation are fused together, and an initial profile representation of the student in terms of knowledge points is obtained through an ability mapping network.

[0089] Optionally, the knowledge point is represented as a vector representation of multiple knowledge points.

[0090] The student's ability representation P and knowledge point representation a are input into a multi-layer neural network-based ability mapping network and fused to output the student's initial profile representation p on the knowledge points.

[0091] p = F([P,a])

[0092] In this embodiment of the invention, the student's ability representation and knowledge point representation are fused together, and an initial profile representation of the student on the knowledge points is obtained through an ability mapping network, thereby achieving accurate prediction of the student profile.

[0093] In some embodiments, the method further includes:

[0094] By combining the error factors and guessing factors in the student's learning process, the initial profile representation is processed to obtain the final profile representation of the student.

[0095] Alternatively, a student may have fully mastered a certain knowledge point during the learning process, but made a mistake while answering a question due to factors such as nervousness or carelessness. In this case, the student's performance during the learning process appears as if they had not mastered the knowledge point. It's understandable that the student's mistake was due to factors such as nervousness or carelessness. A mistake represents the probability of answering incorrectly even when the student has fully mastered the knowledge point. In this situation, the student's performance during the learning process does not completely match their actual mastery, and therefore does not effectively reflect their true ability.

[0096] When students, during the learning process, do not actually master or fully master a certain knowledge point, and answer questions in an ambiguous state, relying on subjective guessing, there is a possibility that the student's guess is correct. In other words, the student's performance during the learning process appears as if they have mastered the knowledge point. Guessing represents the probability that a student answers correctly when they do not fully understand the knowledge point. In this situation, the student's performance during the learning process does not completely match their actual mastery, and therefore cannot effectively reflect the student's true ability.

[0097] The error factors and guessing factors in the student's learning process are initialized to 0, and then optimized and updated as the ability mapping network is trained.

[0098] Because students' performance during the learning process may not be entirely consistent with their actual mastery, in an ambiguous state, students tend to guess or be influenced by mistakes. Combining the error factor s and the guessing factor g in the student's learning process, we obtain the student's final profile representation r, as shown in the following formula:

[0099] r = (1-p)*g + p*(1-s)

[0100] In this embodiment of the invention, by combining the error factors and guessing factors in the student's learning process, the initial profile representation is processed to obtain the final profile representation of the student, which can more accurately and realistically reflect the student's actual mastery of each knowledge point.

[0101] In some embodiments, the process of determining the capability mapping network includes:

[0102] Obtain student ability representation samples and profile representation samples;

[0103] The capability representation samples are input into the initial capability mapping network to obtain the predicted profile representation;

[0104] The initial capability mapping network is obtained by iterating the parameters of the predicted profile representation and the profile representation samples.

[0105] Optionally, the answering process and answering records of students with a certain sample size are used for training to obtain student ability representation samples and profile representation samples.

[0106] The ability representation samples are input into the initial ability mapping network to obtain the student's ability prediction profile representation.

[0107] It should be noted that the student's ability prediction profile is obtained through real training on the student's answer process and answer record, rather than a profile that is randomly assumed and pre-set.

[0108] Based on the difference between the predicted image representation and the image representation sample, the initial capability mapping network is iterated multiple times to fine-tune the initial capability mapping network until the difference between the predicted image representation and the image representation sample is within an acceptable range. Then, the parameter iteration is stopped, and the capability mapping network is obtained.

[0109] In this embodiment of the invention, student ability representation samples and profile representation samples are obtained; the ability representation samples are input into an initial ability mapping network to obtain a predicted profile representation; the parameters of the initial ability mapping network are iterated based on the predicted profile representation and the profile representation samples to obtain the ability mapping network. The ability mapping network realizes the prediction of student profiles and can be applied in scenarios where there are no real student historical information records in learning platforms or smart terminals.

[0110] The student ability profiling device provided in the embodiments of the present invention is described below. The student ability profiling device described below can be referred to in correspondence with the student ability profiling method described above.

[0111] Figure 3This is a schematic diagram of the structure of a student ability profiling device provided in one embodiment of the present invention, as shown below. Figure 3 As shown, the student ability profiling device 300 includes:

[0112] The first acquisition unit 310 models the student's historical answering process to obtain the student's first historical state representation;

[0113] The second acquisition unit 320 is used to acquire the student's second historical state representation based on the student's historical answer records;

[0114] The fusion unit 330 is used to fuse the first historical state representation and the second historical state representation of the student to obtain a fused historical state representation.

[0115] The capability mapping unit 340 is used to perform multi-dimensional capability mapping on the fused historical state representation to obtain the student's capability representation;

[0116] The portrait unit 350 is used to obtain a portrait representation of the student based on the student's ability representation and knowledge point representation.

[0117] In some embodiments, modeling the student's historical answering process to obtain a first historical state representation of the student includes:

[0118] Based on recurrent neural networks, the student's historical answering process is modeled to obtain the student's first historical state representation.

[0119] In some embodiments, obtaining the student's second historical state representation based on the student's historical answer records includes:

[0120] Based on an inductive algorithm, the student's answers to each knowledge point in the historical answer record are statistically analyzed, and the student's historical scores are accumulated in an exponential decay manner to obtain the student's second historical state representation.

[0121] In some embodiments, performing multidimensional capability mapping on the fused historical state representation to obtain the student's capability representation includes:

[0122] Determine the student knowledge matrix;

[0123] Based on the student knowledge matrix, a multidimensional mapping is performed on the fused historical state representation to obtain the student's ability representation.

[0124] In some embodiments, obtaining a student profile representation based on the student's ability representation and knowledge point representation includes:

[0125] The student's ability representation and knowledge point representation are fused together, and an initial profile representation of the student in terms of knowledge points is obtained through an ability mapping network.

[0126] In some embodiments, the student proficiency profiling device 300 further includes:

[0127] The portrait processing unit is used to process the initial portrait representation by combining the error factors and guessing factors in the student's learning process, so as to obtain the final portrait representation of the student.

[0128] In some embodiments, the process of determining the capability mapping network includes:

[0129] Obtain student ability representation samples and profile representation samples;

[0130] The capability representation samples are input into the initial capability mapping network to obtain the predicted profile representation;

[0131] The initial capability mapping network is obtained by iterating the parameters of the predicted profile representation and the profile representation samples.

[0132] It should be noted that the student ability profiling device provided in this embodiment of the invention can implement all the method steps implemented in the above-described student ability profiling method embodiment and can achieve the same technical effect. Therefore, the parts and beneficial effects that are the same as those in the method embodiment will not be described in detail here.

[0133] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a student ability profiling method. This method includes: modeling the student's historical answering process to obtain a first historical state representation of the student; obtaining a second historical state representation of the student based on the student's historical answering records; fusing the first and second historical state representations to obtain a fused historical state representation; performing multi-dimensional ability mapping on the fused historical state representation to obtain the student's ability representation; and obtaining a student profile representation based on the student's ability representation and knowledge point representation.

[0134] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0135] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the student ability profiling method provided in the above-described method embodiments. This method includes: modeling a student's historical answering process to obtain a first historical state representation of the student; obtaining a second historical state representation of the student based on the student's historical answering records; fusing the first and second historical state representations to obtain a fused historical state representation; performing multi-dimensional ability mapping on the fused historical state representation to obtain a student ability representation; and obtaining a student profile representation based on the student's ability representation and knowledge point representation.

[0136] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the student ability profiling method provided in the above-described method embodiments. This method includes: modeling a student's historical answering process to obtain a first historical state representation of the student; obtaining a second historical state representation of the student based on the student's historical answering records; fusing the first and second historical state representations to obtain a fused historical state representation; performing multi-dimensional ability mapping on the fused historical state representation to obtain a student's ability representation; and obtaining a student profile representation based on the student's ability representation and knowledge point representation.

[0137] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0138] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for profiling student abilities, characterized in that, include: Based on recurrent neural networks, the student's historical answering process is modeled to obtain the student's first historical state representation; Based on an inductive algorithm, the student's answers to each knowledge point in the historical answer record are statistically analyzed, and the student's historical scores are accumulated in an exponential decay manner to obtain the student's second historical state representation. By fusing the first and second historical state representations of the student, a fused historical state representation is obtained. Determine the student knowledge matrix; based on the student knowledge matrix, perform multi-dimensional capability mapping on the fused historical state representation to obtain the student's capability representation; Based on the student's ability representation and knowledge point representation, a profile representation of the student is obtained.

2. The student ability profiling method according to claim 1, characterized in that, The process of obtaining a student profile representation based on the student's ability representation and knowledge point representation includes: The student's ability representation and knowledge point representation are fused together, and an initial profile representation of the student in terms of knowledge points is obtained through an ability mapping network.

3. The student ability profiling method according to claim 2, characterized in that, The method further includes: By combining the error factors and guessing factors in the student's learning process, the initial profile representation is processed to obtain the final profile representation of the student.

4. The student ability profiling method according to claim 2 or 3, characterized in that, The process of determining the capability mapping network includes: Obtain student ability representation samples and profile representation samples; The capability representation samples are input into the initial capability mapping network to obtain the predicted profile representation; The initial capability mapping network is obtained by iterating the parameters of the predicted profile representation and the profile representation samples.

5. A student ability profiling device, characterized in that, The first acquisition unit is used to model the student's historical answering process based on a recurrent neural network to obtain the student's first historical state representation; The second acquisition unit is used to statistically analyze the student's answer to each knowledge point in the historical answer record based on an inductive algorithm, and to accumulate the student's historical score in an exponential decay manner to obtain the student's second historical state representation. The fusion unit is used to fuse the student's first historical state representation and second historical state representation to obtain a fused historical state representation. A capability mapping unit is used to determine the student knowledge matrix; based on the student knowledge matrix, multi-dimensional capability mapping is performed on the fused historical state representation to obtain the student's capability representation; A profile unit is used to obtain a profile representation of the student based on the student's ability representation and knowledge point representation.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the student competency profiling method as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the student competency profiling method as described in any one of claims 1 to 4.

Images