Knowledge point tracking method and device, equipment and storage medium

Abstract

The present disclosure relates to a knowledge point tracking method, device, equipment and storage medium. The method comprises: obtaining a answered data sequence corresponding to a historical time period of a student and a to-be-answered data sequence corresponding to a current time of the student, wherein the answered data sequence comprises a historical real answer right or wrong result, a knowledge point representation and a question difficulty representation, and the to-be-answered data sequence comprises a current real knowledge point representation, a question and a question difficulty representation; based on a pre-trained knowledge point tracking model, the answered data sequence and the to-be-answered data sequence are processed to obtain the mastery degree of the knowledge point of the to-be-answered question of the student. Thus, based on the data of multiple dimensions such as the real answering result, the knowledge point and the difficulty of the question of the student, the knowledge point tracking is performed, the differences of the questions under the same knowledge point are considered, the accuracy of the knowledge point tracking result is improved, and the method is suitable for deployment and application in a real education scene.

Description

Technical Field

[0001] This disclosure relates to the field of educational technology, and in particular to a method, apparatus, device, and storage medium for tracking knowledge points. Background Technology

[0002] In the field of intelligent education, in order to provide scalable, data-driven personalized education, it is necessary to track students' mastery of knowledge points in real time. This allows teachers to improve their understanding of students based on the tracking results and provide targeted guidance. It also helps students understand their weaknesses, enabling them to fill in knowledge gaps and avoid the pitfalls of rote memorization, thereby improving students' learning efficiency and quality.

[0003] In existing technologies, to mitigate the impact of data sparsity on model performance, knowledge points are typically used to represent questions. Students' answers to questions are then converted into answers to knowledge points, and these answers are analyzed to predict students' mastery of those knowledge points. However, even questions covering the same knowledge point can differ, making it difficult to capture these differences. This results in poor accuracy in knowledge point tracking, hindering its deployment in real-world educational scenarios. Summary of the Invention

[0004] To address the aforementioned technical problems, this disclosure provides a knowledge point tracking method, apparatus, device, and storage medium.

[0005] Firstly, this disclosure provides a knowledge point tracking method, which includes:

[0006] Obtain the sequence of answered data corresponding to a student in a historical time period and the sequence of unanswered data corresponding to the student in the current time. The sequence of answered data includes the knowledge point representation of the answered questions in multiple historical time periods, the correctness result of the answered questions, the answered questions, and the difficulty representation of the answered questions. The sequence of unanswered data includes the knowledge point representation of the unanswered questions, the unanswered questions, and the difficulty representation of the unanswered questions.

[0007] Based on a pre-trained knowledge point tracking model, the student's mastery of the knowledge points of the questions to be answered is obtained by processing the answered data sequence and the unanswered data sequence.

[0008] Secondly, this disclosure provides a knowledge point tracking device, which includes:

[0009] The acquisition module is used to acquire the sequence of answered data of students in historical time periods and the sequence of data to be answered of students in the current time period. The sequence of answered data includes the knowledge point representation of the answered questions in multiple historical time periods, the correctness result of the answered questions, the answered questions, and the difficulty representation of the answered questions. The sequence of data to be answered includes the knowledge point representation of the questions to be answered, the questions to be answered, and the difficulty representation of the questions to be answered.

[0010] The knowledge point tracking module is used to obtain the student's mastery of the knowledge points of the questions to be answered by processing the answered data sequence and the unanswered data sequence based on a pre-trained knowledge point tracking model.

[0011] Thirdly, embodiments of this disclosure also provide an electronic device, the device comprising:

[0012] processor;

[0013] Memory, used to store executable instructions;

[0014] The processor is used to read executable instructions from memory and execute the executable instructions to implement the method provided in the first aspect above.

[0015] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium having a computer program stored thereon, wherein the storage medium stores the computer program, and when the computer program is executed by a processor, the processor causes the processor to implement the method provided in the first aspect above.

[0016] The technical solution provided in this disclosure has the following advantages compared with the prior art:

[0017] This disclosure discloses a knowledge point tracking method, apparatus, device, and storage medium. It acquires a sequence of answered data from a student over a historical time period and a sequence of data to be answered at the student's current moment. The answered data sequence includes knowledge point representations of multiple answered questions from various historical time periods, the correctness of the answers, the answered questions themselves, and their difficulty levels. The data sequence to be answered includes knowledge point representations of the questions to be answered, the questions themselves, and their difficulty levels. Based on a pre-trained knowledge point tracking model, the method processes the answered and to-be-answered data sequences to determine the student's mastery of the knowledge points for the questions to be answered. Therefore, knowledge point tracking is performed based on multiple dimensions of data, including the student's actual answers, knowledge points, and question difficulty. This process considers the differences in questions covering the same knowledge point, thus achieving accurate knowledge point tracking results and making it suitable for deployment in real-world educational scenarios. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0019] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating a knowledge point tracking method provided in an embodiment of this disclosure;

[0021] Figure 2 A flowchart illustrating another knowledge point tracking method provided in this embodiment of the disclosure;

[0022] Figure 3 A logical schematic diagram of a knowledge point tracking method provided in an embodiment of this disclosure;

[0023] Figure 4 A schematic diagram of the model architecture of a knowledge point tracking model provided in an embodiment of this disclosure;

[0024] Figure 5 This is a schematic diagram of the structure of a knowledge point tracking device provided in an embodiment of the present disclosure;

[0025] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0026] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0027] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0028] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc., used in this disclosure are only used to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0029] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0030] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0031] To improve the accuracy of knowledge point tracking results, this disclosure provides a knowledge point tracking method, apparatus, device, and medium. The following is in conjunction with… Figures 1 to 4 The knowledge point tracking method provided in this disclosure is described below. In this disclosure, the knowledge point tracking method can be executed by an electronic device or a server. Both the electronic device and the server are knowledge point tracking terminals in this disclosure. The electronic device can include devices with communication functions such as mobile phones, tablets, desktop computers, and laptops. The server can be a cloud server or server cluster, or other devices with storage and computing functions. It should be noted that the following embodiments use an electronic device as the execution subject for illustrative explanation.

[0032] Figure 1 A flowchart illustrating a knowledge point tracking method provided in an embodiment of this disclosure is shown.

[0033] like Figure 1 As shown, this knowledge point tracking method may include the following steps.

[0034] S110. Obtain the sequence of answers completed by the student in the historical time period and the sequence of questions to be answered by the student at the current time. The sequence of answers completed includes the knowledge point representation of the questions completed at multiple historical times, the correctness of the answers, the questions completed, and the difficulty representation of the questions completed. The sequence of questions to be answered includes the knowledge point representation of the questions to be answered, the questions to be answered, and the difficulty representation of the questions to be answered.

[0035] In this embodiment, when a student is answering a question, the electronic device can acquire the student's historical answer sequence and current answer sequence. The historical answer sequence and the current answer sequence are the student's actual answer data on the question, so that subsequent knowledge point tracking can be performed based on this actual answer data to determine the student's mastery of the knowledge points of the question being answered.

[0036] Among them, the answered data sequence can be understood as the historical answer sequence, and the data sequence to be answered is the current answer sequence.

[0037] The knowledge point representation of the answered questions refers to the knowledge points involved in the questions that students have already answered. Specifically, the knowledge point representation of the answered questions can be obtained through matrix transformation. Optionally, the knowledge point representation of the answered questions can include the knowledge point representations from time 1 to time t, where the knowledge point representation at time 1 is represented as z. ck 1 The knowledge point representation at time t is denoted as z. ck t .

[0038] The correctness or incorrectness of the answered questions refers to the student's correct or incorrect answer. A correct answer is represented by 1, and an incorrect answer by 0. Specifically, the correctness or incorrectness representation vector can be obtained through matrix transformation, and this vector is used as the correctness or incorrectness result of the answered questions. Optionally, the correctness or incorrectness result of the answered questions can include the correctness or incorrectness results from time 1 to time t, where the correctness or incorrectness result at time 1 is represented as r. qj 1 The correctness of the answer at time t is represented as t_t.

[0039] r qj .

[0040] Here, "answered questions" refers to the questions corresponding to the knowledge points that students have already answered. Specifically, answered questions can be obtained through matrix transformation. Optionally, answered questions may include question v from time 1. ck 1 Problem v up to time t ck t .

[0041] The difficulty representation of the answered questions refers to the difficulty representation vector of the questions related to the knowledge points that the student has answered. Specifically, the difficulty representation of the answered questions can be obtained through matrix transformation. Optionally, the difficulty representation of the answered questions may include the difficulty representation m at time step 1. qj 1 Difficulty representation m up to time t qj t .

[0042] Here, the knowledge point representation of the question to be answered refers to the vector representing the knowledge points involved in the question the student is currently answering. Specifically, the knowledge point representation can be obtained through matrix transformation. Optionally, the knowledge point representation of the question to be answered may include the knowledge point representation z at time t+1. ck t+1 .

[0043] Here, the questions to be answered refer to the questions related to the knowledge points that the student is currently answering. Specifically, the question representation vector of the questions to be answered can be obtained through matrix transformation, and this question representation vector is used as the question to be answered. Optionally, the questions to be answered may include the question v at time t+1. ck t+1 .

[0044] The difficulty representation of the question to be answered refers to the difficulty representation vector of the questions related to the knowledge point the student is currently answering. Specifically, the difficulty representation of the question to be answered can be obtained through matrix transformation. Optionally, the difficulty representation of the question to be answered may include the difficulty representation m at time t+1. qj t+1 .

[0045] S120. Based on the pre-trained knowledge point tracking model, the student's mastery of the knowledge points of the questions to be answered is obtained by processing the answered data sequence and the data sequence to be answered.

[0046] In this embodiment, the electronic device acquires a pre-trained knowledge point tracking model and, based on the pre-trained knowledge point tracking model, processes the answered data sequence and the unanswered data sequence. Specifically, it tracks the knowledge point representation of the answered questions, the correctness of the answered questions, the knowledge point representation of the unanswered questions, the unanswered questions, and the difficulty representation of the unanswered questions. It outputs the student's mastery of the knowledge points of the unanswered questions, thereby taking into account the student's actual answer to the questions and combining the tracking processing with the knowledge points and the difficulty of the questions to improve the accuracy of the knowledge point tracking results.

[0047] Here, mastery level refers to the student's level of understanding of the teacher's guidance involved in answering the questions, which can be represented by a representation vector. Optionally, the representation vector corresponding to the mastery level can be represented as l t+1 .

[0048] Specifically, the level of mastery can serve as a reference for teachers to understand students, and it can also serve as a basis for students to identify and fill in knowledge gaps. This makes it easier for teachers to provide targeted guidance to students, while also helping students understand their weaknesses, ultimately improving students' learning efficiency and quality.

[0049] This disclosure provides a knowledge point tracking method. It acquires a sequence of answered data from a student over a historical time period and a sequence of data to be answered at the student's current moment. The answered data sequence includes knowledge point representations of answered questions from multiple historical time periods, the correctness of the answers, the answered questions themselves, and their difficulty levels. The data sequence to be answered includes knowledge point representations of the questions to be answered, the questions themselves, and their difficulty levels. Based on a pre-trained knowledge point tracking model, the method processes the answered and to-be-answered data sequences to determine the student's mastery of the knowledge points for the questions to be answered. Therefore, knowledge point tracking is performed based on multiple dimensions of data, including the student's actual answers, knowledge points, and question difficulty. This process considers the differences in questions covering the same knowledge point, thus achieving accurate knowledge point tracking results and making it suitable for deployment in real-world educational scenarios.

[0050] In another embodiment of this disclosure, the tracking principle of the pre-trained knowledge point tracking model is refined to facilitate a deeper understanding of the knowledge point tracking process.

[0051] Figure 2 A flowchart illustrating another knowledge point tracking method provided in an embodiment of this disclosure is shown.

[0052] like Figure 2 As shown, this knowledge point tracking method may include the following steps.

[0053] S210. Obtain the sequence of answers completed by the student in the historical time period and the sequence of questions to be answered by the student at the current time. The sequence of answers completed includes the knowledge point representation of the questions completed at multiple historical times, the correctness of the answers, the questions completed, and the difficulty representation of the questions completed. The sequence of questions to be answered includes the knowledge point representation of the questions to be answered, the questions to be answered, and the difficulty representation of the questions to be answered.

[0054] S210 is similar to S110, and will not be described in detail here.

[0055] S220. Based on the knowledge point tracking network in the pre-trained knowledge point tracking model, the student obtains the candidate answers to the questions to be answered by processing the already answered data sequence and the data sequence to be answered.

[0056] In this embodiment, optionally, the knowledge point tracking network includes: a feature extraction subnetwork and a knowledge tracking subnetwork; then, S220 specifically includes the following steps:

[0057] S2201. Based on the feature extraction sub-network, feature fusion processing is performed on the knowledge point representation of the answered questions and the correctness of the answers in the answered questions in the answered data sequence to obtain the fused representation of the answered questions at multiple historical moments.

[0058] S2202. Based on the feature extraction subnetwork, perform psychological measurement processing on the knowledge point representation of the question to be answered, the question to be answered, and the difficulty representation of the question to be answered in the question-to-answer data sequence to obtain the output relationship representation of the question to be answered; and based on the feature extraction subnetwork, perform psychological measurement processing on the knowledge point representation of the question to be answered, the question to be answered, and the difficulty representation of the question to be answered in the answer data sequence to obtain the output relationship representation of the question to be answered.

[0059] S2203. Based on the tracking subnetwork, the fusion representation of the answered questions, the output relationship representation of the questions to be answered, and the output relationship representation of the answered questions are tracked and processed to obtain the student's candidate answers to the questions to be answered.

[0060] Specifically, the feature extraction subnetwork includes: an encoding layer and a relational layer;

[0061] Accordingly, S2201 specifically includes: based on the coding layer in the feature fusion sub-network, performing feature fusion processing on the knowledge point representation of the answered questions and the correctness of the answers to the answered questions, to obtain the fused representation of the answers at multiple historical moments.

[0062] Specifically, the knowledge point representation of answered questions includes the knowledge point representation z of answered questions. ck 1 To z ck t The correctness of the answers to the questions already answered includes r. qj 1 to r qj t The coding layer includes 1 to n coding modules, and the first coding module fuses z ck 1 and r qj 1 The first time step yields the fused representation y1 of the responses, and the second encoding module fuses z. ck 2 and r qj 2 The fused representation y2 of the response at time 1 is obtained, and so on, the fused representation z of the encoding module at time t is obtained. ck t and r qj t The historical response interaction representation vector is obtained and used as the fused response representation y at time t. t This yields the fused representation of responses from multiple historical moments, i.e., y1 to y2.t .

[0063] Optionally, the fused representation y of the responses at time t t It can be determined in the following way:

[0064] y t =z ck t +r qj t

[0065] Therefore, the coding layer can be used to fuse historical data sequences to obtain a fused representation of answers from multiple historical moments, which facilitates subsequent tracking of knowledge points using the fused representation of answers from multiple historical moments.

[0066] Accordingly, in S2202, “psychological measurement processing is performed on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the data sequence to be answered based on the feature extraction sub-network to obtain the output relation representation of the question to be answered” specifically includes: based on the relation layer in the feature extraction sub-network, psychological measurement processing is performed on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered to obtain the output relation representation of the question to be answered;

[0067] Meanwhile, in S2202, "based on the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation, the answered questions, and the difficulty of the answered questions in the answered data sequence to obtain the output relationship representation of the answered questions," specifically includes: based on the relationship layer in the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation, the answered questions, and the difficulty representation of the answered questions to obtain the output relationship representation of the answered questions.

[0068] Among them, the relation layer learns a psychometric model (such as Rasch) to construct the relationship between knowledge points and questions. It can perform psychometric processing on knowledge points and questions to obtain the joint representation vector between questions and knowledge points, that is, to obtain the output relation representation of questions to be answered and the output relation representation of questions already answered.

[0069] Specifically, the knowledge points of the questions to be answered are represented by z. ck t+1 The question to be answered is indicated by v. ck t+1 The difficulty level of the questions to be answered is represented by m. qj t+1 This indicates that the relation layer is related to z. ck t+1 v ck t+1 and m qj t+1Psychological measurement processing was performed to obtain the output relation representation x of the questions to be answered. t+1 .

[0070] Optionally, the output relation representation x of the question to be answered t+1 It can be determined in the following way:

[0071] x t+1 =z ck t+1 +m qj t+1 *v ck t+1

[0072] Similarly, the knowledge point representation of the answered questions includes z ck 1 To z ck t The questions already answered include v ck 1 to v ck t The difficulty level of the answered questions includes m qj 1 to m qj t Then the relation layer starts with t=1, for z ck 1 v ck 1 and m qj 1 Psychological measurement processing was performed to obtain the output relation representation x of the first answered question. 1 Until the relation layer reaches z ck t v ck t and m qj t Perform joint processing to obtain the output relation representation x of the t-th answered question. t .

[0073] Optionally, the output relation representation x of the question to be answered t It can be determined in the following way:

[0074] x t =z ck t +m qj t *v ck t

[0075] Therefore, the feature extraction sub-network in the pre-trained knowledge point tracking model can process the historical data sequence and the current data sequence separately. Specifically, the encoding layer is used to fuse the historical data sequence, and the relation layer is used to perform psychological measurement processing on the knowledge points and questions of the current data sequence and the historical data respectively, so as to obtain the differences in questions with the same knowledge points.

[0076] The tracking subnetwork further includes: a self-attention layer and a prediction layer;

[0077] Accordingly, S2203 specifically includes: based on the self-attention layer in the tracking sub-network, performing self-attention processing on the fusion representation of the answered questions, the output relationship representation of the questions to be answered, and the output relationship representation of the answered questions to obtain a self-attention answer representation; based on the prediction layer in the tracking sub-network, performing tracking processing on the self-attention answer representation and the output relationship representation of the questions to be answered to obtain candidate answer results.

[0078] Specifically, the self-attention layer can use a dot product-based attention calculation method to fuse the responses from y1 to y2. t The output relationship representation of the questions to be answered is x. t+1 And the output relationship representation x of the already answered questions t Self-attention processing is performed to obtain the self-attention response representation h. t Furthermore, the prediction layer can further fuse the self-attention response representation h. t The output relationship between x and the question to be answered is represented by x. t+1 The study tracks students' understanding of the knowledge points represented in their answers to questions, and obtains candidate answers. t+1 .

[0079] Among them, the self-attention response representation h t It can be represented in the form of a self-attention hidden vector.

[0080] Optionally, candidate response results l t+1 It can be determined in the following way:

[0081] l t+1 =f(W1*[h t x t+1 ]+b1)

[0082] Where W1 and b1 are constants.

[0083] Therefore, the tracking subnetwork in the pre-trained knowledge point tracking model can track the fusion representation of the answered questions, the output relationship representation of the questions to be answered, and the output relationship representation of the answered questions, thereby obtaining the predicted answer result as a candidate answer result.

[0084] S230. Based on the candidate responses, determine whether the response is correct or incorrect.

[0085] In this embodiment, optionally, S230 specifically includes:

[0086] Based on the interpretability prediction network in the pre-trained knowledge point tracking model, the candidate answers of students are processed to obtain the correctness of the answers.

[0087] Specifically, S230 includes:

[0088] Obtain the probability of guessing correctly and the probability of carelessness for the questions to be answered;

[0089] Based on an interpretable prediction network, abnormal correct answers are processed by considering candidate responses, probability of guessing correctly, and probability of carelessness, thus obtaining the correct or incorrect answer result.

[0090] Here, the probability of guessing correctly refers to the likelihood that a student will guess the correct answer to a question. Optionally, the probability of guessing correctly can be expressed as p. t+1 g express.

[0091] Here, the probability of carelessness refers to the student's prediction of the likelihood of making a careless mistake when answering a question. Optionally, the probability of carelessness can be expressed as p. t+1 s express.

[0092] Understandably, to improve the likelihood of students tracking knowledge points, it is also necessary to comprehensively consider the possibility that students might answer the questions by chance or carelessly, and to predict whether the students' answers to the questions are correct or not. For this reason, this embodiment can utilize an interpretability prediction network to process candidate answers, the probability of guessing correctly, and the probability of carelessness to obtain an answer that excludes the possibility of abnormal correct answers, which is then used as the correct answer result a. t+1 .

[0093] Optionally, indicate the correct or incorrect answer (a). t+1 It can be determined in the following way:

[0094] a t+1 =l t+1 *(1-p t+1 s )+(1-l t+1 )p t+1 g

[0095] Therefore, after determining candidate answers using a pre-trained knowledge point tracking model, the model comprehensively considers factors such as carelessness and guessing to predict student performance and arrives at correct or incorrect answers. This eliminates abnormal correct answers, provides a reasonable explanation for the prediction results of the knowledge point tracking model, and further improves the accuracy of the knowledge point tracking results.

[0096] To facilitate understanding the knowledge point tracking process, Figure 3 A logical diagram of the knowledge point tracking method is shown. Figure 4 A schematic diagram of the model architecture for the knowledge point tracking model is shown.

[0097] like Figure 3 and Figure 4 As shown, this knowledge point tracking method includes the following steps:

[0098] S1. Obtain the knowledge point representation of the answered questions. ck 1 To z ck t The correctness of the answers to the questions already answered (r) qj 1 to r qj t Answered questions v ck 1 to v ck t Difficulty level of already answered questions (m) qj 1 to m qj t The knowledge points represented by the questions to be answered (z) ck t +1 Questions to be answered (v) ck t+1 And the difficulty level of the questions to be answered (m) qj t+1 .

[0099] S2. Using the encoding layer, the knowledge points of the answered questions are represented. ck 1 To z ck t The correctness of the answers to the questions already answered (r) qj 1 to r qj t Feature fusion processing is performed to obtain the fused representations y1 to y2 of the responses. t .

[0100] S3. Utilizing relational layer representations of knowledge points for questions to be answered. ck t+1 Questions to be answered (v)ck t+1 And the difficulty level of the questions to be answered (m) qj t+1 Psychological measurement processing was performed to obtain the output relation representation x of the questions to be answered. t+1 And the use of relational layers to represent the knowledge points of already answered questions. ck 1 To z ck t Answered questions v ck 1 to v ck t Difficulty level of already answered questions (m) qj 1 to m qj t Psychological measurement processing was performed to obtain the output relation representation x of the answered questions. 1 To x t .

[0101] S4. Utilize the self-attention layer to fuse the responses from y1 to y2. t The output relationship representation of the questions to be answered is x. t+1 The output relationship representation of the already answered questions is x. 1 To x t Self-attention processing is performed to obtain the self-attention response representation h. t .

[0102] S5. Utilizing the prediction layer to represent self-attention responses h t The output relationship between x and the question to be answered is represented by x. t+1 The tracking process is performed to obtain the candidate response results. t+1 .

[0103] S6. Get the questions to be answered (v) ck t+1 The corresponding probability p of guessing correctly t+1 g And the probability of carelessness p t+1 s .

[0104] S7. Using an interpretable prediction network, analyze the candidate responses. t+1 The probability of guessing correctly is p. t+1 g And the probability of carelessness p t+1 s Perform error handling to obtain the correct / incorrect answer result a. t+1 .

[0105] This disclosure also provides a knowledge point tracking device for implementing the above-described knowledge point tracking method, which will be described below in conjunction with... Figure 5The following explanation is provided. In this embodiment, the knowledge point tracking device can be an electronic device or a server. The electronic device can include devices with communication functions such as mobile phones, tablets, desktop computers, and laptops. The server can be a cloud server or server cluster, or other devices with storage and computing functions.

[0106] Figure 5 A schematic diagram of the structure of a knowledge point tracking device provided in an embodiment of this disclosure is shown.

[0107] like Figure 5 As shown, the knowledge point tracking device 500 may include:

[0108] The acquisition module 510 is used to acquire the sequence of answered data corresponding to the student in a historical time period and the sequence of data to be answered corresponding to the student in the current time. The sequence of answered data includes the knowledge point representation of the answered questions in multiple historical time periods, the correctness result of the answered questions, the answered questions, and the difficulty representation of the answered questions. The sequence of data to be answered includes the knowledge point representation of the questions to be answered, the questions to be answered, and the difficulty representation of the questions to be answered.

[0109] The knowledge point tracking module 520 is used to obtain the student's mastery of the knowledge points of the questions to be answered by processing the answered data sequence and the question-to-answer data sequence based on a pre-trained knowledge point tracking model.

[0110] This disclosure provides a knowledge point tracking device that acquires a sequence of answered data from a student over a historical time period and a sequence of data to be answered from a student at the current time. The answered data sequence includes knowledge point representations of answered questions from multiple historical time periods, the correctness of the answers, the answered questions themselves, and their difficulty levels. The data sequence to be answered includes knowledge point representations of the questions to be answered, the questions themselves, and their difficulty levels. Based on a pre-trained knowledge point tracking model, the device processes the answered and to-be-answered data sequences to determine the student's level of mastery of the knowledge point representations for the questions to be answered. Therefore, knowledge point tracking is performed based on multiple dimensions of data, including the student's actual answers, knowledge points, and question difficulty. This process considers the differences in questions covering the same knowledge point, thereby achieving accurate knowledge point tracking results and making it suitable for deployment in real-world educational scenarios.

[0111] In some optional embodiments, the knowledge point tracking module 520 includes:

[0112] The processing module is used to process the answered data sequence and the unanswered data sequence based on the knowledge point tracking network in the pre-trained knowledge point tracking model to obtain the student's candidate answer results for the unanswered question;

[0113] The determination module is used to determine the correctness of the answer based on the candidate answer results.

[0114] In some optional embodiments, the knowledge point tracking network includes: a feature extraction subnetwork and a tracking subnetwork; then, the processing module includes:

[0115] The first processing unit is used to perform feature fusion processing on the knowledge point representation of the answered questions and the correctness result of the answered questions in the answered data sequence based on the feature extraction sub-network, so as to obtain the fused representation of the answered questions at multiple historical moments.

[0116] The second processing unit is used to perform psychological measurement processing on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the question-to-answer data sequence based on the feature extraction sub-network to obtain the output relationship representation of the question to be answered; and to perform psychological measurement processing on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the answered data sequence based on the feature extraction sub-network to obtain the output relationship representation of the question to be answered.

[0117] The third processing unit is used to perform tracking processing on the already answered fusion representation, the output relationship representation of the question to be answered, and the output relationship representation of the already answered question based on the tracking sub-network, so as to obtain the student's candidate answer result for the question to be answered.

[0118] In some optional embodiments, the first processing unit is specifically used for:

[0119] Based on the encoding layer in the feature fusion subnetwork, feature fusion processing is performed on the knowledge point representation of the answered questions and the correctness of the answers to the answered questions to obtain the fused representation of the answers at multiple historical moments.

[0120] In some optional embodiments, the second processing unit is specifically used for:

[0121] Based on the relation layer in the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation of the question to be answered, the question to be answered, and the difficulty representation of the question to be answered to obtain the output relation representation of the question to be answered.

[0122] Based on the relation layer in the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation of the answered questions, the answered questions, and the difficulty representation of the answered questions to obtain the output relation representation of the answered questions.

[0123] In some optional embodiments, the third processing unit is specifically used for:

[0124] Based on the self-attention layer in the tracking sub-network, self-attention processing is performed on the already answered fusion representation, the output relationship representation of the question to be answered, and the output relationship representation of the already answered question to obtain a self-attention answer representation;

[0125] Based on the prediction layer in the tracking subnetwork, the self-attention answer representation and the output relationship representation of the question to be answered are tracked to obtain the candidate answer result.

[0126] In some alternative embodiments, the determining module includes:

[0127] The acquisition unit is used to acquire the probability of guessing correctly and the probability of carelessness corresponding to the question to be answered;

[0128] The fourth processing unit is used to perform abnormal answer processing on the candidate answer results, the probability of guessing correctly and the probability of carelessness based on the interpretability prediction network, so as to obtain the answer correctness result.

[0129] It should be noted that, Figure 5 The knowledge point tracking device 500 shown can perform... Figures 1 to 4 The various steps in the method embodiment shown are implemented. Figures 1 to 4 The various processes and effects in the method embodiments shown are not elaborated here.

[0130] Exemplary embodiments of this disclosure also provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, which, when executed by the at least one processor, causes the electronic device to perform a method according to an embodiment of this disclosure.

[0131] Exemplary embodiments of this disclosure also provide a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform a method according to embodiments of this disclosure.

[0132] Exemplary embodiments of this disclosure also provide a computer program product, including a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform a method according to an embodiment of this disclosure.

[0133] refer to Figure 6 The present invention describes a structural block diagram of an electronic device 600 that can serve as a server or client of the present disclosure. This is an example of a hardware device that can be applied to various aspects of the present disclosure, and the electronic device 600 can be the aforementioned electronic device. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0134] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0135] Multiple components in electronic device 600 are connected to I / O interface 605, including: input unit 606, output unit 607, storage unit 608, and communication unit 609. Input unit 606 can be any type of device capable of inputting information to electronic device 600. Input unit 606 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of electronic device. Output unit 607 can be any type of device capable of presenting information and may include, but is not limited to, a display, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 608 may include, but is not limited to, disks and optical discs. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers, and / or chipsets, such as Bluetooth™ devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0136] The computing unit 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above. For example, in some embodiments, the knowledge point tracking method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 600 via ROM 602 and / or communication unit 609. In some embodiments, the computing unit 601 can be configured to perform the knowledge point tracking method by any other suitable means (e.g., by means of firmware).

[0137] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0138] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0139] As used in this disclosure, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0140] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0141] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0142] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0143] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0144] The above are merely specific embodiments of this disclosure, enabling those skilled in the art to understand or implement this disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to these embodiments, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A knowledge point tracking method, characterized in that, include: Obtain the sequence of answered data corresponding to a student in a historical time period and the sequence of unanswered data corresponding to the student in the current time. The sequence of answered data includes the knowledge point representation of the answered questions in multiple historical time periods, the correctness result of the answered questions, the answered questions, and the difficulty representation of the answered questions. The sequence of unanswered data includes the knowledge point representation of the unanswered questions, the unanswered questions, and the difficulty representation of the unanswered questions. Based on a pre-trained knowledge point tracking model, the student's candidate answers to the question to be answered are obtained by processing the already answered data sequence and the question to be answered. Then, based on the interpretability prediction network in the pre-trained knowledge point tracking model, the student's candidate answers are processed to obtain the correctness of the answer to the question to be answered. The step of obtaining the student's candidate answers to the question to be answered includes: The knowledge point tracking network in the pre-trained knowledge point tracking model includes: a feature extraction subnetwork and a tracking subnetwork. Based on the feature extraction subnetwork, feature fusion processing is performed on the knowledge point representation of the answered questions and the correctness of the answers in the answered data sequence to obtain the fused representation of the answered questions at multiple historical moments. Based on the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the question-to-answer data sequence to obtain the output relationship representation of the question to be answered; and based on the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the answered data sequence to obtain the output relationship representation of the question to be answered. Based on the tracking subnetwork, the student's candidate answers to the questions to be answered are obtained by tracking the fusion representation of the answers, the output relationship representation of the questions to be answered, and the output relationship representation of the answers.

2. The method according to claim 1, characterized in that, The feature extraction subnetwork performs feature fusion processing on the knowledge point representations of the answered questions and the correctness results of the answered questions in the answered data sequence to obtain fused representations of the answered questions at multiple historical moments, including: Based on the encoding layer in the feature extraction subnetwork, feature fusion processing is performed on the knowledge point representation of the answered questions and the correctness of the answers to the answered questions to obtain the fused representation of the answers at multiple historical moments.

3. The method according to claim 1, characterized in that, The feature extraction subnetwork performs psychometric processing on the knowledge point representations, the questions to be answered, and the difficulty representations of the questions to be answered in the data sequence to be answered, to obtain the output relationship representation of the questions to be answered, including: Based on the relation layer in the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation of the question to be answered, the question to be answered, and the difficulty representation of the question to be answered to obtain the output relation representation of the question to be answered. The feature extraction subnetwork performs psychometric processing on the knowledge point representations, answered questions, and difficulty levels of the answered questions in the answered data sequence to obtain the output relationship representation of the answered questions, including: Based on the relation layer in the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation of the answered questions, the answered questions, and the difficulty representation of the answered questions to obtain the output relation representation of the answered questions.

4. The method according to claim 1, characterized in that, The step of tracking the fused representation of the answered questions, the output relationship representation of the questions to be answered, and the output relationship representation of the answered questions based on the tracking subnetwork to obtain the student's candidate answers to the questions to be answered includes: Based on the self-attention layer in the tracking sub-network, self-attention processing is performed on the already answered fusion representation, the output relationship representation of the question to be answered, and the output relationship representation of the already answered question to obtain a self-attention answer representation; Based on the prediction layer in the tracking subnetwork, the self-attention answer representation and the output relationship representation of the question to be answered are tracked to obtain the candidate answer result.

5. The method according to claim 1, characterized in that, The interpretability prediction network in the pre-trained knowledge point tracking model is used to process the student's candidate answers to obtain the correctness of the answer, including: Obtain the probability of guessing correctly and the probability of carelessness for the question to be answered; Based on the interpretability prediction network, abnormal answer processing is performed on the candidate answer results, the probability of guessing correctly, and the probability of carelessness to obtain the answer correctness result.

6. A knowledge point tracking device, characterized in that, include: The acquisition module is used to acquire the sequence of answered data of students in historical time periods and the sequence of data to be answered of students in the current time period. The sequence of answered data includes the knowledge point representation of the answered questions in multiple historical time periods, the correctness result of the answered questions, the answered questions, and the difficulty representation of the answered questions. The sequence of data to be answered includes the knowledge point representation of the questions to be answered, the questions to be answered, and the difficulty representation of the questions to be answered. The knowledge point tracking module is used to process the answered data sequence and the unanswered data sequence based on a pre-trained knowledge point tracking model to obtain the student's candidate answers to the unanswered questions. Then, based on the interpretability prediction network in the pre-trained knowledge point tracking model, the module processes the student's candidate answers to obtain the correctness of the answer to the unanswered questions. The step of obtaining the student's candidate answers to the question to be answered includes: The knowledge point tracking network in the pre-trained knowledge point tracking model includes: a feature extraction subnetwork and a tracking subnetwork. Based on the feature extraction subnetwork, feature fusion processing is performed on the knowledge point representation of the answered questions and the correctness of the answers in the answered data sequence to obtain the fused representation of the answered questions at multiple historical moments. Based on the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the question-to-answer data sequence to obtain the output relationship representation of the question to be answered; and based on the feature extraction subnetwork, psychological measurement processing is performed on the knowledge point representation, the question to be answered, and the difficulty representation of the question to be answered in the answered data sequence to obtain the output relationship representation of the question to be answered. Based on the tracking subnetwork, the student's candidate answers to the questions to be answered are obtained by tracking the fusion representation of the answers, the output relationship representation of the questions to be answered, and the output relationship representation of the answers.

7. An electronic device, characterized in that, include: processor; Memory, used to store executable instructions; The processor is configured to read the executable instructions from the memory and execute the executable instructions to implement the method of any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, The storage medium stores a computer program that, when executed by a processor, causes the processor to implement the method described in any one of claims 1-5.

Images