Automatic question generation method, question recommendation method, electronic equipment and storage medium

By extracting core knowledge points from electronic textbooks and generating prompt words, and entering large models to generate questions and standard answers, the problem of low-quality question generation efficiency in the existing technology is solved, and automatic automatic generation and recommendation of questions is realized, and efficiency is improved.

CN119003753APending Publication Date: 2024-11-22WU HAN XIN ZHI SHU ZI KE JI YOU XIAN GONG SI
View PDF 0 Cites 5 Cited by

Patent Information

Application Number
CN202411210193.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In the prior art, the generation efficiency of the question is low, mainly relying on manual generation, resulting in inefficiency.

Method used

By extracting core knowledge points from electronic textbooks, generating corresponding prompt words, and inputting them into a big model to generate questions and standard answers, including checksum correction of the accuracy of standard answers, and using keywords to generate classification tags, recommending users Matching questions.

Benefits of technology

It improves the efficiency of question generation, solves the problem of low manual generation efficiency, and realizes automatic question generation and recommendation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119003753A_ABST
    Figure CN119003753A_ABST
Patent Text Reader

Abstract

The invention provides an automatic question generation method, a question recommendation method, electronic equipment and a storage medium. The automatic question generation method comprises the steps that core knowledge points are extracted from an electronic teaching material, then corresponding cue words are generated for the core knowledge points, then the core knowledge points and the cue words are input into a large model, and at the moment, due to the fact that the cue words comprise the types of questions needing to be generated and question generation modes, the types of the questions needing to be generated and the types of the questions needing to be generated can be generated. Therefore, the questions and the standard answers corresponding to the questions can be generated through the large model. Compared with an existing manual question generation mode, the question generation mode obviously can improve the question generation efficiency, and therefore the problems in the prior art are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to a method for identifying user emotions, an electronic device, and a storage medium. Background Art

[0002] With the development of science and technology, more and more people are learning online through terminal devices such as computers, mobile phones, and tablets. In the process of online learning, it is obviously essential to obtain questions to do and test.

[0003] Currently, questions are mainly generated manually, for example, by experienced teaching staff editing questions based on their teaching experience, but this method of generating questions is relatively inefficient. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a user emotion recognition method, a topic recommendation method, an electronic device and a storage medium to solve the problems in the prior art.

[0005] A first aspect of an embodiment of the present application provides a method for automatically generating a topic, comprising:

[0006] Extract core knowledge points from electronic textbooks;

[0007] Generate corresponding prompt words for the core knowledge points; the prompt words include the type of questions to be generated and the question generation method;

[0008] The core knowledge points and the prompt words are input into the big model to generate questions and the standard answers corresponding to the questions.

[0009] Preferably, the method further comprises:

[0010] Verifying the accuracy of the standard answer;

[0011] In the case of failing the verification, the standard answer is manually corrected and replaced by the corrected standard answer.

[0012] Preferably, the method further comprises:

[0013] Extracting keywords from the standard answer corresponding to the question;

[0014] The keyword is used to generate a classification label for the topic.

[0015] Preferably, generating corresponding prompt words for the core knowledge points specifically includes:

[0016] Obtaining a prompt word generation template, wherein the prompt word generation template includes an execution flow for indicating generation of questions and standard answers;

[0017] According to the execution process in the prompt word generation template, the prompt words corresponding to the core knowledge points are generated.

[0018] Preferably, generating prompt words corresponding to the core knowledge points according to the execution flow in the prompt word generation template specifically includes:

[0019] Determine the importance level of the core knowledge points;

[0020] Determining the type of questions to be generated according to the importance level;

[0021] Obtain a question generation method corresponding to the type;

[0022] The prompt word is generated according to the type and the question generation method.

[0023] Preferably, the types specifically include fill-in-the-blank questions, multiple-choice questions, true-or-false questions and short-answer questions; and,

[0024] The method of generating questions corresponding to the fill-in-the-blank questions specifically includes: deleting at least one of time, person, place, and event from the core knowledge points;

[0025] The question generation method corresponding to the multiple-choice question specifically includes: extracting at least one of time, person, place, and event from the core knowledge point, and setting options according to the extracted information;

[0026] The question generation method corresponding to the judgment question specifically includes: setting the test point of the judgment question for at least one of the time, person, place, and event in the core knowledge point;

[0027] The question generation method corresponding to the short-answer question specifically includes: generating a section of explanation content for the core knowledge point, and setting corresponding short-answer questions for the explanation content, wherein the explanation content is used to provide a general introduction to the core knowledge point.

[0028] The second aspect of the embodiment of the present application provides a method for recommending questions, wherein the question bank includes multiple questions and standard answers and classification labels corresponding to each question, and the questions are generated by the method provided by the first aspect of the present application, and the method includes:

[0029] Get user tags;

[0030] Determine the similarity between the user tag and the classification tags of each question in the question bank;

[0031] According to the similarity, a target question is selected from each question in the question bank for recommendation to the user, wherein the target question is specifically a question whose similarity between the classification label and the user label is greater than a preset threshold, or the N questions with the largest similarity between the classification label and the user label, wherein N is an integer greater than or equal to 1.

[0032] Preferably, obtaining the user tag specifically includes:

[0033] The user tag is generated through user behavior data and user attribute information.

[0034] A third aspect of the embodiments of the present application provides an electronic device, including:

[0035] Memory for storing computer programs;

[0036] A processor is used to execute any one of the methods described in the method embodiments of the present application.

[0037] A fourth aspect of an embodiment of the present application provides a storage medium, including: a program, which, when executed on an electronic device, enables the electronic device to execute any one of the methods in the method embodiments of the present application.

[0038] The automatic question generation method provided by the embodiment of the present application includes extracting core knowledge points from the electronic textbook, generating corresponding prompt words for the core knowledge points, and then inputting the core knowledge points and the prompt words into the large model. At this time, since the prompt words include the type of question to be generated and the question generation method, the question and the standard answer corresponding to the question can be generated through the large model. Compared with the current manual question generation method, this question generation method can obviously improve the efficiency of question generation, thus solving the problems in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application;

[0041] Figure 2 A schematic diagram of a specific flow chart of a method for automatically generating a topic provided in an embodiment of the present application;

[0042] Figure 3A schematic diagram of a specific flow chart of a method for recommending a topic provided in an embodiment of the present application;

[0043] Figure 4 A schematic diagram of the specific structure of an automatic topic generation device provided in one embodiment of the present application;

[0044] Figure 5 A schematic diagram of the specific structure of a topic recommendation device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings in the embodiments of the present application. In the description of the present application, terms such as "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance or sequence.

[0046] As mentioned above, with the development of science and technology, more and more people are learning online. In the process of online learning, it is obviously necessary to obtain questions for practice and testing. Currently, questions are mainly generated manually, but the efficiency of generating questions in this way is relatively low.

[0047] In view of this, the embodiments of the present application provide a method for automatically generating a topic, a method for recommending a topic, a device, an electronic device, and a storage medium, which can be used to solve the technical problem. Figure 1 FIG. 1 is a schematic diagram of a specific structure of an electronic device provided in this embodiment. The electronic device 1 includes: at least one processor 11 and a memory 12. Figure 1 A processor is used as an example. The processor 11 and the memory 12 may be connected via a bus 10, and the memory 12 stores instructions that can be executed by the processor 11, and the instructions are executed by the processor 11, so that the electronic device 1 can execute all or part of the process of the method in the embodiment of the present application.

[0048] Among them, the electronic device 1 can be a client electronic device (referred to as a client), such as a mobile phone, a computer, etc., and the electronic device 1 can also be a server electronic device (referred to as a server), such as a server node of the server, etc.

[0049] like Figure 2 The figure is a schematic diagram of a specific process of the user emotion recognition method provided by the embodiment of the present application. The method can be performed by Figure 1 The method is performed by the electronic device 1 shown in the figure. Here, the method is described by taking the electronic device 1 as a server to perform the method. In this case, the method includes the following steps:

[0050] Step S21: extracting core knowledge points from the electronic textbook.

[0051] The electronic teaching material may be a teaching material in electronic form, and certain knowledge points need to be extracted from the electronic teaching material to generate questions. In practical applications, the electronic teaching material may be directly generated through a paper teaching material, for example, the electronic teaching material may be generated by scanning a paper teaching material. It should be emphasized that the electronic teaching material may also be an electronic teaching material obtained through secondary processing, for example, relevant teaching staff processes and extracts knowledge from teaching syllabuses, textbooks, etc., and then produces electronic files such as PPT documents, videos, and audios for teaching. At this time, the electronic files such as PPT documents, videos, and audios for teaching may be used as the electronic teaching materials in step S21.

[0052] Among them, the core knowledge points in step S21 are knowledge points that can be used to set questions. For example, various knowledge points are usually included in the teaching syllabus. These knowledge points can be used as core knowledge points for subsequent generation of questions. Or, according to the importance level of these knowledge points, some important knowledge points can be used as the core knowledge points for subsequent generation of questions.

[0053] In practical applications, there are many ways to implement step S21. For example, the electronic textbook can be converted into text first, and then the core knowledge points in the text can be extracted using the knowledge point extraction function of large models such as GPT (Generative Pre-Trained) and Wenxinyiyan. Specifically, for example, the core knowledge points can be converted into text first through relevant software, such as using speech recognition software to recognize electronic files in audio or video form to generate text, and then the text can be input into the large model as a whole or in blocks, so as to use the knowledge point extraction function of the large model to extract the core knowledge points in the text as the core knowledge points extracted from the electronic textbook.

[0054] In practical applications, one method is to input the entire text into a large model, thereby extracting core knowledge points through the large model. However, this method is likely to affect the accuracy of the extracted core knowledge points when the text is too long, such as when facing a long text. At this time, the text can usually be divided into blocks to obtain multiple sub-texts. The length of these sub-texts needs to be lower than a preset threshold. In this way, each sub-text can be input into the large model separately to obtain the core knowledge points of each sub-text. At this time, the core knowledge points of each sub-text can be used as the core knowledge points of the electronic textbook.

[0055] Step S22: Generate corresponding prompt words for the core knowledge point.

[0056] Among them, the prompt word contains restrictions on the questions to be generated, such as the restrictions used to limit the types of questions to be generated (such as fill-in-the-blank questions, multiple-choice questions, true-or-false questions, short-answer questions, etc.), limit the method of generating questions (i.e., the method of generating questions), etc. Therefore, for the core knowledge points, the prompt word can be used to generate corresponding questions.

[0057] There may be multiple ways to implement step S22 in actual applications. For example, one way is to first obtain a prompt word generation template, which includes an execution process for indicating the generation of questions and standard answers, and then generate the prompt word corresponding to the core knowledge point according to the execution process in the prompt word generation template.

[0058] For example, the execution process includes first determining the importance level of the core knowledge point. At this time, the examination outline can be queried to verify the importance level of the core knowledge point, and then the type of question to be generated is determined based on the importance level. For example, if the importance level of the core knowledge point is low, the type can usually be fill-in-the-blank question, multiple-choice question or true-or-false question. If the importance level of the core knowledge point is high, the type can usually be short-answer question.

[0059] After obtaining the type of question, obtain the question generation method corresponding to the type. For example, if the type is a fill-in-the-blank question, the corresponding question generation method may be, for example: delete at least one of time, person, place, and event from the core knowledge point, so that the question (i.e., fill-in-the-blank question) can be generated according to the question generation method; if the type is a multiple-choice question, the corresponding question generation method may be, for example: first extract at least one of time, person, place, and event from the core knowledge point, and then set options according to the extracted information. For example, if time is extracted, the time can be used as one of the options, and at least one interference item can be further set to obtain the generated question (i.e., multiple-choice question).

[0060] If the type is a true or false question, the corresponding question generation method may be, for example: setting the test points of the true or false question for at least one of the time, person, place, event and other information in the core knowledge point, such as modifying at least one of the time, person, place, event and other information to generate the question (i.e., the true or false question); if the type is a short answer question, the corresponding question generation method may be, for example: generating a section of explanation content for the core knowledge point, which may be a general introduction to the core knowledge point, and then setting corresponding short answer questions for the explanation content, such as the short answer questions may be "how do you view the incident", "what are the subsequent impacts of the incident", "how do you evaluate the historical figure", etc., to generate the question (i.e., the short answer question).

[0061] In this way, through the above method, the type of questions to be generated and the method of generating the questions to be generated can be obtained for the core knowledge points, so that prompt words can be generated according to the type and the method of generating the questions. For example, the type of question and the method of generating the question can be used as the corresponding prompt words for generating the core knowledge points.

[0062] It should be noted that, in actual applications, for the specific method of obtaining the prompt word generation template, a general prompt word generation template can usually be generated in advance, so that the general prompt word generation template can be directly obtained, or the prompt word generation template corresponding to the knowledge point can be generated for different core knowledge points. Among them, the general prompt word generation template usually contains a general execution process, which is used to indicate the generation of questions and standard answers.

[0063] Step S23: Input the core knowledge point and the prompt word into the large model to generate a question and a standard answer corresponding to the question.

[0064] After obtaining the prompt word corresponding to the core knowledge point through the above-mentioned step S22, in this step S23, the core knowledge point and the prompt word are further input into the big model. Since the prompt word includes the type of question to be generated and the question generation method, the big model can generate questions for the core knowledge point according to the prompt word. Of course, while generating the questions, the standard answers corresponding to the questions can also be generated.

[0065] The method provided by the embodiment of the present application includes extracting core knowledge points from the electronic textbook, generating corresponding prompt words for the core knowledge points, and then inputting the core knowledge points and the prompt words into the large model. At this time, since the prompt words include the type of questions to be generated and the question generation method, the large model can be used to generate questions and the standard answers corresponding to the questions. Compared with the current method of manually generating questions, this question generation method can obviously improve the efficiency of generating questions, thus solving the problems in the prior art.

[0066] Therefore, in practical applications, electronic files such as PPT documents, videos, and audios for teaching can be obtained, and then these electronic files can be converted into text through relevant software, and then the core knowledge points in the text can be extracted using functions such as GPT knowledge point extraction. After obtaining the core knowledge point, a prompt word generation template can be obtained, and then the prompt word corresponding to the core knowledge point is generated according to the execution flow in the prompt word generation template. After generating the prompt word corresponding to the core knowledge point, the core knowledge point and the prompt word can be further input into the large model to generate the question and the standard answer corresponding to the question.

[0067] It should be emphasized that since the present application generates questions and standard answers through a large model, considering that the reliability of the current large model may be insufficient, it is often necessary to verify the accuracy of the standard answers. Therefore, the present application may also include verification of the accuracy of the standard answers when generating questions and the standard answers corresponding to the questions. For example, the accuracy of the standard answers may be verified manually, or the standard answers may be substituted into the questions, and then both are input into the large model again, so as to determine through the large model whether the standard answers accurately answer the questions in the questions, thereby verifying the accuracy of the standard answers.

[0068] At this time, if the standard answer passes the verification, it means that the accuracy of the standard answer is relatively high, so it can be ignored; on the contrary, if it fails the verification, it means that the accuracy of the standard answer is relatively low. At this time, the standard answer can be manually corrected and replaced with the corrected standard answer, that is, the corrected standard answer can be used as the standard answer corresponding to the question.

[0069] After obtaining the title and the standard answer corresponding to the title, it is usually necessary to classify the title to facilitate its management. Therefore, the method can further include extracting the keywords in the standard answer corresponding to the title, and then using the keywords to generate the classification label of the title, so as to classify the title with the classification label. Therefore, in an embodiment of the present application, the classification label of the title is generated by the keywords in the standard answer, so as to achieve the classification of the title. Among them, the keywords in the standard answer can usually be information such as time, person, place, event, etc., so that after obtaining the keywords, these keywords can be directly used as the classification label, and the keywords can also be further classified, such as classifying multiple times into one period, etc.

[0070] In the embodiment of the present application, after generating questions and standard answers by the above method, the questions and standard answers can be stored in a question bank, wherein the question bank includes multiple questions and the standard answers and classification labels corresponding to these questions. In practical applications, in order to improve the utilization rate of questions in the question bank, it is usually necessary to recommend the questions in the question bank to users. Therefore, the questions generated by the question generation method provided in the embodiment of the present application can further provide a method for recommending questions, such as Figure 3 The figure shows a specific flow chart of the topic recommendation method, which includes the following steps:

[0071] Step S31: Obtain user tags.

[0072] The user tag is used to reflect the classification of the user. For example, the user tag can reflect information such as the user's gender, age, education level, test-taking habits, shopping habits, etc., thereby reflecting the classification of the user.

[0073] In practical applications, the specific implementation of step S31 may be, for example, to generate the user tag through user behavior data and user attribute information, wherein the user behavior data reflects the user's historical behavior, and the user attribute data reflects the user's attributes. For example, the user behavior data may include information on dynamic behaviors such as the user's test-taking habits and shopping habits; and the user attribute data may include information on relatively static attributes such as the user's gender, age, and education level.

[0074] In an embodiment of the present application, with the user's authorization, the user's behavior data can be obtained, such as the user's test-taking habits, shopping habits, etc., and the user's attribute data can also be obtained, such as the user's gender, age, education level, etc., so as to combine the two to generate a user tag for the user.

[0075] Step S32: Determine the similarity between the user tag and the classification tags of each question in the question bank.

[0076] Step S33: According to the similarity, a target question is selected from each question in the question bank and recommended to the user.

[0077] Here, the step S32 and the step S33 can be uniformly explained. After the user tag is obtained through the above step S31, in the step S32, the similarity between the user tag and the classification tags of each question in the question bank can be further determined.

[0078] For example, the user label can be converted into a vector (called the first vector), and the classification labels of each question in the question bank can also be converted into vectors (called the second vector). Then, the similarity between the first vector and each second vector is calculated. The similarity can be cosine similarity, so as to obtain the similarity between the user label and the classification labels of each question in the question bank. At this time, the higher the similarity, the more matched they are, and vice versa.

[0079] After obtaining the similarity between the user tag and the classification tags of each question in the question bank, target questions can be selected from the questions in the question bank for recommendation to the user based on the similarity, wherein the target questions are specifically questions whose similarity between the classification label and the user tag is greater than a preset threshold, or the N questions with the largest similarity ranking between the classification label and the user label, wherein N is an integer greater than or equal to 1.

[0080] For example, a preset threshold can be set, and then based on the similarity between the user tag and the classification tags of each question in the question bank, the question with a similarity greater than the preset threshold can be selected as the target question to be recommended to the user.

[0081] Of course, the similarities between the user tags and the classification tags of each question in the question bank can also be sorted from large to small or from small to large, and then the N questions with the largest similarity rankings are selected as the target questions to be recommended to the user. For example, N can be 10, 15, 20 or other values. In this way, this method can recommend questions that are more suitable for users to improve their usage experience.

[0082] Of course, in the question recommendation method, since the questions in the question bank are generated by the question generation method provided by the embodiment of the present application, it can also solve the problems in the prior art, which will not be elaborated here.

[0083] It should also be noted that, in the embodiment itself, the questions in the question bank are generated by the question map generation method provided by the embodiment of the present application, and in the automatic generation method of the question map, step S23 is to input the core knowledge point and the prompt word into the large model to generate the question and the standard answer corresponding to the question, and the current large model may have various problems of insufficient performance or unknown, such as the generated question is too bizarre or does not meet the actual needs of the user, so the method can further include using the user's evaluation information on each question, the evaluation information includes negative evaluation information, positive evaluation information and neutral evaluation information, etc., these evaluation information are used to negatively, positively and neutrally evaluate the corresponding question, respectively. After collecting the user's evaluation information on each question, the proportion of negative evaluation information of each question can be determined. For example, for a certain question, the number of negative evaluation information of the question is in the total number of evaluation information of the question, so as to obtain the proportion of negative evaluation information of the question, and then for the question whose proportion of negative evaluation information is greater than the upper limit value, it means that the question does not meet the actual needs of the user. At this time, the question can be deleted from the question bank, and the core knowledge point corresponding to the question is used to regenerate the question and the standard answer.

[0084] Based on the same inventive concept as the method for automatically generating questions provided in the embodiment of the present application, the embodiment of the present application also provides a device for automatically generating questions. If there are any unclear points, please refer to the corresponding content of the method embodiment. Figure 4 The specific structural diagram of the device 40 is shown, and the device 40 may include: an acquisition unit 401, an extraction unit 402 and an identification unit 403, wherein:

[0085] Extraction unit 401, used to extract core knowledge points from the electronic textbook;

[0086] A prompt word generating unit 402 is used to generate corresponding prompt words for the core knowledge points; the prompt words include the type of question to be generated and the question generation method;

[0087] The question generating unit 403 is used to input the core knowledge points and the prompt words into the large model to generate questions and the standard answers corresponding to the questions.

[0088] The device 40 provided in the embodiment of the present application is adopted. Since the device 40 adopts the same inventive concept as the method for automatically generating questions provided in the embodiment of the present application, on the premise that the method can solve the technical problem, the device 40 can also solve the technical problem, which will not be repeated here.

[0089] In practical applications, the device 40 may also include a verification and correction unit for verifying the accuracy of the standard answer; if the standard answer fails the verification, the standard answer is corrected manually and replaced with the corrected standard answer.

[0090] The device 40 may also include a question classification unit, which is used to extract keywords from the standard answer corresponding to the question; and generate a classification label for the question using the keywords.

[0091] The generating of corresponding prompt words for the core knowledge points may specifically include:

[0092] Obtaining a prompt word generation template, wherein the prompt word generation template includes an execution flow for indicating generation of questions and standard answers;

[0093] According to the execution process in the prompt word generation template, the prompt words corresponding to the core knowledge points are generated.

[0094] According to the execution flow in the prompt word generation template, generating the prompt word corresponding to the core knowledge point may specifically include:

[0095] Determine the importance level of the core knowledge points;

[0096] Determining the type of questions to be generated according to the importance level;

[0097] Obtain a question generation method corresponding to the type;

[0098] The prompt word is generated according to the type and the question generation method.

[0099] The types specifically include fill-in-the-blank questions, multiple-choice questions, true-or-false questions, and short-answer questions; and,

[0100] The method of generating questions corresponding to the fill-in-the-blank questions specifically includes: deleting at least one of time, person, place, and event from the core knowledge points;

[0101] The question generation method corresponding to the multiple-choice question specifically includes: extracting at least one of time, person, place, and event from the core knowledge point, and setting options according to the extracted information;

[0102] The question generation method corresponding to the judgment question specifically includes: setting the test point of the judgment question for at least one of the time, person, place, and event in the core knowledge point;

[0103] The question generation method corresponding to the short-answer question specifically includes: generating a section of explanation content for the core knowledge point, and setting corresponding short-answer questions for the explanation content, wherein the explanation content is used to provide a general introduction to the core knowledge point.

[0104] Based on the same inventive concept as the method for recommending questions provided in the embodiment of the present application, the embodiment of the present application also provides a device for recommending questions. If there is any unclear point, please refer to the corresponding content of the method embodiment. The question bank includes multiple questions and the standard answers and classification labels corresponding to each question. The questions are generated by the method for generating questions provided in the embodiment of the present application. Figure 5 FIG. 5 is a schematic diagram of a specific structure of the device 50. The device 50 may include: a user tag acquisition unit 501, a similarity determination unit 502, and a recommendation unit 503, wherein:

[0105] A user tag acquisition unit 501, used to acquire a user tag;

[0106] A similarity determination unit 502 is used to determine the similarity between the user tag and the classification tag of each question in the question bank;

[0107] The recommendation unit 503 is used to select a target question from each question in the question bank and recommend it to the user according to the similarity, wherein the target question is specifically a question whose similarity between the classification label and the user label is greater than a preset threshold, or the N questions with the largest similarity between the classification label and the user label, wherein N is an integer greater than or equal to 1.

[0108] Obviously, the device 50 can also solve the problems in the prior art, which will not be described in detail here.

[0109] Wherein, obtaining the user tag may specifically include: generating the user tag through user behavior data and user attribute information.

[0110] The embodiment of the present invention further provides a storage medium, including: a program, which, when running on an electronic device, enables the electronic device to execute all or part of the process of the method in the above embodiment. The storage medium may be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk (HDD) or a solid-state drive (SSD). The storage medium may also include a combination of the above types of memories.

[0111] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for automatically generating a topic, characterized in that: include: Extract core knowledge points from electronic textbooks; Generate corresponding prompt words for the core knowledge points; The prompt words include the type of question to be generated and the question generation method; The core knowledge points and the prompt words are input into the big model to generate questions and the standard answers corresponding to the questions.

2. The method according to claim 1, characterized in that The method further comprises: Verifying the accuracy of the standard answer; In the case of failing the verification, the standard answer is manually corrected and replaced by the corrected standard answer.

3. The method according to claim 1, characterized in that The method further comprises: Extracting keywords from the standard answer corresponding to the question; The keyword is used to generate a classification label for the topic.

4. The method according to claim 1, characterized in that: Generate corresponding prompt words for the core knowledge points, specifically including: Obtaining a prompt word generation template, wherein the prompt word generation template includes an execution flow for indicating generation of questions and standard answers; According to the execution process in the prompt word generation template, the prompt words corresponding to the core knowledge points are generated.

5. The method according to claim 4, characterized in that Generating prompt words corresponding to the core knowledge points according to the execution process in the prompt word generation template specifically includes: Determine the importance level of the core knowledge points; Determining the type of questions to be generated according to the importance level; Obtain a question generation method corresponding to the type; The prompt word is generated according to the type and the question generation method.

6. The method according to claim 5, characterized in that The types specifically include fill-in-the-blank questions, multiple-choice questions, true-or-false questions, and short-answer questions; and, The method of generating questions corresponding to the fill-in-the-blank questions specifically includes: deleting at least one of time, person, place, and event from the core knowledge points; The question generation method corresponding to the multiple-choice question specifically includes: extracting at least one of time, person, place, and event from the core knowledge point, and setting options according to the extracted information; The question generation method corresponding to the judgment question specifically includes: setting the test point of the judgment question for at least one of the time, person, place, and event in the core knowledge point; The question generation method corresponding to the short-answer question specifically includes: generating a section of explanation content for the core knowledge point, and setting corresponding short-answer questions for the explanation content, wherein the explanation content is used to provide a general introduction to the core knowledge point.

7. A method for recommending a topic, characterized in that: The question bank includes a plurality of questions and standard answers and classification labels corresponding to each question, wherein the questions are generated by the method according to any one of claims 1 to 6, and the method includes: Get user tags; Determine the similarity between the user tag and the classification tags of each question in the question bank; According to the similarity, a target question is selected from each question in the question bank for recommendation to the user, wherein the target question is specifically a question whose similarity between the classification label and the user label is greater than a preset threshold, or the N questions with the largest similarity between the classification label and the user label, wherein N is an integer greater than or equal to 1.

8. The method according to claim 7, characterized in that Obtaining user tags specifically includes: The user tag is generated through user behavior data and user attribute information.

9. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to execute the method according to any one of claims 1 to 8.

10. A storage medium, characterized in that: The invention comprises: a program, which, when running on an electronic device, enables the electronic device to execute the method as claimed in any one of claims 1 to 8.

Citation Information

Cited By

  • Teaching material generation method and device, electronic equipment and program product

    CN119293197A

  • A teaching material generation method, device, electronic equipment and program product

    CN119293197B

  • Question bank construction method and system, storage medium and electronic equipment

    CN119358512A

  • Intelligent English question setting method and system based on large model

    CN119850378A

  • Intelligent proposition method and device for examination, equipment and storage medium

    CN121257748A