Test paper composition method and device, electronic equipment, storage medium and computer program product

Abstract

The invention provides a test paper composition method and device, electronic equipment, a storage medium and a computer program product. The method comprises the following steps: receiving a test paper composition instruction of a user; performing intention analysis on the test paper composition instruction to obtain a test paper composition intention of the user; and customizing personalized test paper of the user based on the test paper composition intention and the portrait information of the user. The objective portrait information and the subjective test paper composition intention of the user are fully considered, so that the test paper better meeting the personalized requirements of the user can be generated, the learning efficiency and the learning effect of the user are improved, and the problems that the test paper content is not matched with the user requirements, the pertinence is not high and the like in a traditional test paper composition method are solved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a test paper assembling method, device, electronic device, storage medium and computer program product. Background Art

[0002] With the continuous development of educational informatization, the demand for personalized learning is growing. Test-taking, as a crucial step in the teaching process, directly impacts student learning outcomes. Providing efficient and accurate personalized test-taking services for diverse student groups is a pressing issue in education.

[0003] At present, the common method of test paper composition mainly relies on teachers to manually compose the test papers. Although this method can combine teaching experience, the scope of test paper composition is often based on a wide range of general rules, which makes it difficult to take into account the personalized learning needs of each student. As a result, the test paper composition results lack specificity and cannot effectively meet the needs of individual students such as strengthening their weaknesses. Summary of the Invention

[0004] The present invention provides a test paper assembling method, device, electronic device, storage medium and computer program product to solve the defects in the prior art.

[0005] The present invention provides a test paper composing method, comprising the following steps: Receive the user's test preparation instruction; Performing intention analysis on the test-taking instruction to obtain the test-taking intention of the user; Based on the test paper composition intention and the user's portrait information, a personalized test paper for the user is customized.

[0006] According to a test paper composition method provided by the present invention, the method of customizing a personalized test paper for a user based on the test paper composition intention and the user's portrait information includes: Based on the portrait information, determining multiple test type combinations that match the user; Determine the recommendation weight of each question under each test type based on the test intention, wherein the test intention is used to represent the test requirements of the user in different dimensions; The test paper is compiled based on the recommended weight of each test question under each test paper type to obtain a personalized test paper for the user.

[0007] According to a test paper composition method provided by the present invention, the step of determining multiple test paper composition types matching the user based on the portrait information includes: Based on the portrait information, multiple test paper types that match the user are determined, as well as the difficulty ratio of knowledge points under each test paper type.

[0008] According to a test paper composition method provided by the present invention, determining the weight of each test question under each test paper composition type based on the test paper composition intention includes: Based on the difficulty ratio of knowledge points in each test paper type, determine the first recommendation weight of each test question in each test paper type; Based on the test paper setting intention, determining the second recommendation weight of each test question under each test paper setting type; The first recommendation weight and the second recommendation weight of each test question are integrated to obtain the recommendation weight of each test question.

[0009] According to a test paper composition method provided by the present invention, determining the second recommendation weight of each test question under each test paper composition type based on the test paper composition intention includes: Determine the initial recommended weight of each question based on the knowledge point type of each question under each test paper type; Based on the test paper composition intention, the initial weight of each test question is adjusted to obtain a second recommended weight for each test question.

[0010] According to a test paper composition method provided by the present invention, the test paper composition type carries a priority tag; the test paper composition is performed based on the recommended weight of each test question under each test paper composition type to obtain the user's personalized test paper, including: Based on the recommended weights of each question under each test paper type, the test paper is compiled to obtain personalized test papers corresponding to each test paper type; Based on the priority labels of each test paper type, the personalized test papers corresponding to each test paper type are prioritized.

[0011] According to a test paper composition method provided by the present invention, the test paper is composed based on the recommended weight of each test question under each test paper composition type to obtain the user's personalized test paper, and then further includes: receiving a tuning instruction from the user; Based on the tuning information carried by the tuning instruction, optimizing the recommended weight of each test question under each test paper type; The test paper is compiled based on the optimized recommended weights of each test question to obtain an optimized personalized test paper.

[0012] According to a test paper composition method provided by the present invention, performing intention analysis on the test paper composition instruction to obtain the user's test paper composition intention includes: Based on the portrait information, determining multiple candidate test composition intentions; Based on the roll formation instruction, the roll formation intention of the user is determined from the multiple candidate roll formation intentions.

[0013] The present invention also provides a test paper assembly device, comprising the following modules: A receiving unit, configured to receive a test paper generating instruction from a user; An analysis unit, configured to perform an intention analysis on the test-taking instruction to obtain the test-taking intention of the user; The customization unit is used to customize the user's personalized test paper based on the test paper composition intention and the user's portrait information.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the above-described test paper assembly methods is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements any of the above-mentioned methods for assembling test papers when executed by a processor.

[0016] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned test paper assembling methods.

[0017] The test paper-generating method, device, electronic device, storage medium, and computer program product provided by the present invention analyze the user's test paper-generating instructions to determine the user's test paper-generating intention, and combine this with the user's profile information to ultimately generate a personalized test paper. Because the present invention fully considers the user's objective profile information and subjective test paper-generating intention, it can generate a test paper that better meets the user's personalized needs, improving the user's learning efficiency and learning outcomes, and avoiding the problems of traditional test paper-generating methods such as the mismatch between test paper content and user needs and lack of targetedness. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 This is one of the flow charts of the test paper composition method provided by the present invention.

[0020] Figure 2 It is a flowchart of the implementation of step 130 in the test paper composition method provided by the present invention.

[0021] Figure 3 It is a flowchart of the implementation of step 132 in the test paper composition method provided by the present invention.

[0022] Figure 4 This is the second flow chart of the test paper assembling method provided by the present invention.

[0023] Figure 5 This is the third flow chart of the test paper assembling method provided by the present invention.

[0024] Figure 6 This is the fourth flow chart of the test paper assembling method provided by the present invention.

[0025] Figure 7 It is a structural schematic diagram of the paper assembly device provided by the present invention.

[0026] Figure 8 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0027] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0028] At present, common test-taking methods mainly include the following two: one is to rely on teachers to manually compose the test papers. Although this method can combine teaching experience, the scope of test-taking is often based on a wide range of general rules, which makes it difficult to take into account the personalized learning needs of each student. As a result, the test-taking results lack pertinence and cannot effectively meet the needs of individual students such as strengthening their weaknesses. The second is test-taking products based on fixed rules. Such products on the market often use fixed dimensions or question banks. After the user makes a selection, they can only generate specific results. There is a lack of personalized test-taking processes and results, and it is difficult to dynamically adjust and optimize according to the student's own situation. At the same time, the accuracy of identifying user intentions needs to be improved, and it is impossible to truly understand the real needs of students. Therefore, traditional test-taking methods have obvious deficiencies in personalization and intelligence, and cannot meet the growing demand for personalized learning.

[0029] To this end, the present invention provides a test paper composition method. Figure 1 This is one of the flow charts of the test paper composition method provided by the present invention, such as Figure 1 As shown, the method includes step 110 , step 120 and step 130 .

[0030] Step 110: Receive the user's test paper generation instruction.

[0031] Here, the term "user" refers to an individual who needs to be assessed using a generated test paper. This individual can be a student, a school, or an educational district, and this is not specifically limited in this embodiment of the present invention. A test paper generation instruction refers to instruction information used to express a user's test paper generation needs. It is used to characterize the characteristics and requirements that the user wishes to generate a test paper, such as difficulty, question type, knowledge point coverage, and emphasis.

[0032] Among them, the test paper generation instructions may include: ① Voice instructions, which refer to the test paper generation requirements expressed by the user through voice input. For example, the user can say "I want a physics test paper that focuses on the law of conservation of momentum, has a moderate difficulty level, and includes multiple-choice questions and calculation questions." ② Text instructions, which refer to text information entered by the user through a keyboard, touch screen, etc., used to describe their specific requirements for the test paper in detail. For example, the user can enter "Generate a test paper on electromagnetic induction with a medium difficulty level, including multiple-choice questions, fill-in-the-blank questions, and calculation questions, focusing on the application of Faraday's law of electromagnetic induction" in a dialog box. ③ Preset function button instructions, which refer to the user triggering the generation of a test paper with specific attributes by clicking a specific function button preset on the learning device. For example, the learning device can have function buttons such as "Redo Wrong Questions," "Reinforce Knowledge Points," and "Mock Exam" for the user to choose from.

[0033] Taking into account that there may be ambiguity or incompleteness when users express their intentions, after receiving the user's test paper generation instructions, the instructions can be preprocessed first, such as text cleaning, stop word removal, etc. Then, the large language model can be used to semantically understand and complete the instructions, and the vague instructions can be converted into clear and unambiguous intention descriptions. Finally, the instructions can be supplemented or adjusted based on the user's historical learning data. For example, if the user only raised the need for "midterm review" in the test paper generation instructions, specific knowledge points, question types, difficulty and other information can be automatically supplemented based on the user's midterm exam scope, historical wrong question records and knowledge point mastery, to generate a test paper that better meets the user's personalized needs, so as to better meet the user's personalized needs, improve the accuracy and efficiency of subsequent intent analysis, and ensure the pertinence and effectiveness of the test paper generation results.

[0034] In addition, to further enhance the user experience, pre-set test-setting instruction templates or prompts can be provided to guide users in more clearly expressing their test-setting requirements. For example, a template such as "Please generate a test paper of [difficulty], [subject], focusing on [knowledge point]" can be provided for users to refer to and modify.

[0035] Step 120: Perform intent analysis on the test paper generation instruction to obtain the user's test paper generation intention.

[0036] Specifically, test-setting intent refers to the user's expectations and goals for the test, representing the assessment objectives and content scope that the user hopes to achieve through the test. Test-setting intent is derived through intent analysis of test-setting instructions. For example, analyzing the natural language instruction "Generate a 12th grade mechanics final review paper" indicates the user's intent is to generate a 12th grade mechanics final review paper.

[0037] Among them, the test paper creation intention often includes the user's test paper creation needs in multiple different dimensions, such as the difficulty of knowledge points, the attributes of knowledge points (including frequently tested points, easy-to-make mistakes, and weak points), the local test situation, the question type and other dimensions. For example, under the knowledge point difficulty dimension, users can choose difficulty levels such as "easy", "medium", and "difficult" to control the overall difficulty of the test paper. Under the knowledge point attribute dimension, users can choose attributes such as "frequently tested points", "easy-to-make mistakes", and "weak points" to highlight the key content of the test paper. Under the local test situation dimension, users can choose "local high-frequency test points" to make the test paper more in line with local test trends. Under the question type dimension, users can choose question types such as "multiple-choice questions", "fill-in-the-blank questions", and "answer questions" to meet different assessment needs.

[0038] As an optional embodiment, a large language model can be used to perform in-depth semantic analysis on the test-taking instructions, extract the keywords and key information contained in the test-taking instructions, and then match these keywords and key information with the pre-built knowledge graph to determine the user's test-taking intention, such as the range of knowledge points and difficulty level that the user wants to be tested. Furthermore, the user's historical learning data and behavioral habits can be combined to make more accurate predictions and judgments about the user's test-taking intention. For example, if the user often makes mistakes on a certain knowledge point, it can be inferred that the user wants to generate a test paper that focuses on testing that knowledge point.

[0039] Furthermore, to improve the accuracy of intent analysis, we can use multiple rounds of dialogue to interact with users and guide them to more clearly express their test-setting requirements. For example, if we cannot accurately understand the user's input instructions, we can proactively ask the user, "What knowledge points do you want the test to focus on?" This will help us more accurately understand the user's test-setting intentions.

[0040] Step 130: Customize the user's personalized test paper based on the test paper composition intention and the user's profile information.

[0041] Specifically, user profile information refers to a description of user characteristics, which is used to represent information such as the user's learning level, knowledge mastery, and learning preferences. User profile information may include information such as the user's grade, subject, history learning background, weak points, frequently tested points, and study time.

[0042] Methods for determining the profile information may include: ① Obtaining the user's basic information and learning status by having the user actively complete a questionnaire. ② Constructing the user's knowledge graph by analyzing the user's historical learning data, such as homework completion, test scores, and wrong answer records. ③ Inferring the user's learning interests and preferences by combining the user's social information and behavioral data, such as the activity level in the learning community and search history. It is understood that other methods can also be used to determine the profile information, and the embodiments of the present invention do not specifically limit this.

[0043] Unlike traditional methods, this embodiment of the present invention not only considers the user's objective profile information when generating personalized test papers, but more importantly, fully incorporates the user's subjective intentions in creating the test paper. This approach gives users greater autonomy and flexibility, allowing the generated test paper to truly reflect the user's will and needs, achieving a personalized learning experience that is "different for each individual."

[0044] Specifically, when generating a test paper, you can first determine the test paper type and the difficulty ratio of knowledge points based on the user's portrait information. For example, if the user's historical academic situation shows that he is in the mid-term review stage and there are many wrong questions in the mechanics section, you can give priority to mid-term review type test papers and appropriately increase the proportion of medium-difficulty questions in the mechanics section. Then, you can make fine adjustments to the test paper based on the user's test paper composition intention. For example, if the user expresses through the test paper composition instruction that "I want a test paper that focuses on the law of conservation of energy and contains some application questions related to real life", you can give priority to questions related to the law of conservation of energy on the premise of meeting the mid-term review and mechanics ratios, and tend to choose application questions that are combined with real life to improve the user's learning interest and problem-solving ability.

[0045] In addition, in order to further enhance the personalization of the test paper, the test paper's layout style, font size, question presentation method and other details can be adjusted according to the user's learning preferences. For example, if the user is accustomed to using a tablet computer for learning, a test paper suitable for tablet reading can be generated, and pictures and animations can be appropriately added to the questions to improve the user's reading experience.

[0046] In summary, the embodiment of the present invention combines the user's objective portrait information with the subjective test paper composition intention to generate a highly personalized test paper to meet the user's different learning needs and preferences, thereby improving the user's learning efficiency and learning effect.

[0047] For example, user A's profile indicates they are a student with weak foundations, and in their test-taking instructions, they specify, "I want a test paper that's easier and includes detailed explanations to help me consolidate my basic knowledge." Based on this, a test paper with lower difficulty can be generated based on user A's profile. Detailed explanations are provided after each question, helping user A understand the thinking and approach to solving the questions.

[0048] The test-taking method in this embodiment of the present invention is more like an intelligent agent, capable of understanding the user's needs and dynamically adjusting and optimizing the test content based on the user's individual characteristics, generating a personalized test paper that truly meets the user's expectations. Compared to traditional semi-customized template-based test-taking methods, this invention is more flexible and adaptable, better reflects the student's will, and can effectively improve students' learning efficiency and interest.

[0049] It should be noted that the above steps 120 and 130 can be respectively implemented through a large language model (hereinafter referred to as the "big model"). For example, in step 120, the big model can be used to analyze the test paper composing instructions input by the user and extract the user's test paper composing intention; in step 130, the big model can be used to select appropriate questions from the question bank based on the user's test paper composing intention and portrait information, and combine and sort them to finally generate a personalized test paper. Steps 120 and 130 can also be implemented at one time through the big model, such as inputting the user's portrait information and test paper composing instructions into the big model, and the big model directly generates a complete test paper. The embodiment of the present invention does not specifically limit this. The big model here refers to a deep learning model that is trained based on massive data and has powerful natural language processing and generation capabilities. It can be a Transformer model, a BERT model, etc.

[0050] The test paper-generating method provided by an embodiment of the present invention analyzes the user's test-generating instructions to determine the user's test-generating intentions, and combines this with the user's profile information to ultimately generate a personalized test paper. Because this embodiment fully considers the user's objective profile information and subjective test-generating intentions, it can generate test papers that better meet the user's personalized needs, improving their learning efficiency and effectiveness. It also avoids the problems of traditional test-generating methods, such as the mismatch between test paper content and user needs and a lack of targetedness.

[0051] Based on the above embodiments, Figure 2 is a flow chart of an embodiment of step 130 provided by the present invention, as shown in FIG. Figure 2 As shown, step 130 includes: Step 131: Based on the portrait information, determine multiple test paper types that match the user.

[0052] Specifically, the test paper type refers to the overall positioning and format of the test paper, which is used to indicate the design direction and focus of the test paper. Among them, the test paper type can apply to standard test papers (such as standard papers, advanced papers, etc.), specialized practice papers (such as papers for individual weaknesses, papers for local common mistakes, papers for local high-frequency test points, etc.), simulation papers (such as middle school entrance examination simulation papers, college entrance examination simulation papers, etc.), and test paper packages (such as basic consolidation test packages, advanced training test packages, etc.). For example, standard papers focus on assessing students' mastery of basic knowledge, while advanced papers focus on assessing students' comprehensive application ability. Practice papers for individual weaknesses provide intensive training on students' weak points, while practice papers for local common mistakes focus on practicing common mistakes made by local students.

[0053] Considering that the portrait information contains the user's grade, subject, historical grades, weak knowledge points, frequently tested knowledge points, study time and other information, on this basis, the relevant examination paper types can be preliminarily screened out according to the user's grade and subject; then, the examination paper types can be fine-tuned according to the user's historical grades and weak knowledge points.

[0054] For example, if the user's portrait information shows that the user is a sophomore in high school, has poor grades in mechanics, and has weaknesses in the knowledge point of "Kinetic Energy Theorem", you can recommend the "High School Mechanics - Kinetic Energy Theorem - Personal Weakness Points Training Paper".

[0055] For example, if the user's profile information shows that the user is a third-year junior high school student with average math grades, has recently performed poorly in the "quadratic function" section, and has studied for a short time, then the test paper types that can be recommended to him include: "Third-year junior high school mathematics - quadratic function - basic consolidation question package" and "Third-year junior high school mathematics - quadratic function - standard test paper".

[0056] As an optional embodiment, candidate test paper types can be first screened out from a predefined test paper type library based on the user's grade and subject; then, the matching degree between the user portrait information and each candidate test paper type is calculated, and the matching degree can be calculated based on factors such as the user's historical grades, weak knowledge points, and learning time; then, the candidate test paper types are sorted according to the matching degree, and several test paper types with the highest matching degree are selected as the test paper types that match the user; finally, these test paper types can be appropriately adjusted and optimized, for example, the weights of weak knowledge points can be adjusted according to the user's latest learning situation, and finally multiple test paper types that match the user are obtained.

[0057] It is understood that after obtaining multiple test-taking types, the multiple test-taking types can be displayed (e.g., in a list or card format) to encourage the user to select the test-taking type that most interests or requires them. If the user selects a test-taking type, the test-taking process for that type can be directly initiated. If the user is not satisfied with any of the test-taking types, the user can be guided to manually enter their test-taking requirements, or the user can be provided with the option to customize the test-taking type.

[0058] Step 132: Based on the test paper setting intention, determine the recommendation weight of each test question under each test paper setting type. The test paper setting intention is used to represent the user's test paper setting needs in different dimensions.

[0059] Specifically, the recommendation weight of each test question refers to the probability of the test question being selected into the final test paper under a specific test paper type. It is used to characterize the degree of match between the test question and the user's test paper intention and its contribution to improving the quality of the test paper. The greater the recommendation weight, the more the test question conforms to the user's test paper intention, and the higher the priority of the corresponding test paper recommended to the user as a test paper result. For example, for a test paper type such as "Senior Year Mechanics-Kinetic Energy Theorem-Personal Weaknesses", test questions that examine knowledge points related to the Kinetic Energy Theorem and are of moderate difficulty can be given a higher recommendation weight.

[0060] In addition, the test paper intention includes the test paper needs of the user in multiple different dimensions. The test paper needs in different dimensions refer to the user's specific requirements for the test paper in different aspects such as knowledge points, difficulty, question types, and test conditions, such as the test paper needs in dimensions such as knowledge point difficulty, knowledge point attributes, local test conditions, and question types. For example, the user may hope that the test paper is of moderate difficulty, focuses on the application of the kinetic energy theorem, and contains a certain proportion of multiple-choice questions and essay questions. Based on this, the embodiment of the present invention will comprehensively consider the user's test paper needs in various dimensions, and combine the characteristics of the test questions themselves (such as the knowledge points, difficulty, question types, local test conditions, etc.), calculate the recommended weight of each test question, and finally obtain the recommended weight of each test question under each test paper type.

[0061] For example, if the user specifies the test requirements of "medium difficulty", "examining the kinetic energy theorem", and "containing 50% multiple-choice questions and 50% essay questions", then for a multiple-choice question that examines the application of the kinetic energy theorem, if its difficulty is rated as medium and it is in line with the local test situation, its recommendation weight will be relatively high; conversely, if the difficulty of the question is too high or too low, or it examines other knowledge points, its recommendation weight will be relatively low.

[0062] As an optional embodiment, you can first filter out candidate test questions related to the knowledge points specified by the user; then, calculate the scores of each candidate test question in terms of difficulty, question type, local test situation and other dimensions, and the calculation of the scores can be based on predefined rules or machine learning models; then, add up the scores in each dimension and weight them to obtain a comprehensive score for each candidate test question, which is the recommended weight of the test question; finally, normalize the recommended weights of all test questions so that their values ​​are between 0 and 1, and finally obtain the recommended weights of each test question under each test paper type.

[0063] It is understandable that before determining the recommended weight of each test question, the test questions in the question bank can be pre-processed, for example, the test questions can be labeled with knowledge points, rated for difficulty, classified into question types, analyzed for local test conditions, etc., so that the subsequent calculation of the recommended weight can be more efficient and accurate.

[0064] Step 133: Compose a test paper based on the recommended weights of each test question under each test paper type to obtain a personalized test paper for the user.

[0065] Specifically, the recommendation weight of each question is used to indicate how well it matches the user's needs and how much it contributes to improving the quality of the test paper. This weight can then be used to sort and filter the questions, selecting those with higher recommendation weights and combining them into a complete test paper for the user. Because this test paper is generated based on the user's personalized profile information and the test-setting intent, it can better meet the user's learning needs, more specifically strengthen their weak points, and improve their learning efficiency, resulting in a more personalized test paper and better learning outcomes.

[0066] As an optional embodiment, all test questions can be first sorted in descending order according to the recommended weight of each test question; then, from the sorted test question list, test questions are selected in turn to be added to the test paper until the preset number of questions or total score requirements of the test paper are met; in the process of selecting test questions, the knowledge point relevance and difficulty matching between the test questions need to be considered to ensure the overall quality of the test paper and obtain the user's test paper.

[0067] As another optional embodiment, a test paper composition method based on an optimization algorithm can also be adopted. For example, the test paper composition problem can be modeled as an integer programming problem, where the objective function is to maximize the sum of the recommended weights of all test questions in the test paper, and the constraints include the number of questions in the test paper, the total score, the knowledge point coverage, the difficulty distribution, etc., and then an optimization algorithm (such as linear programming, genetic algorithm, etc.) is used to solve the integer programming problem to obtain the optimal test paper and the user's test paper.

[0068] The test paper composition method provided by the embodiment of the present invention can generate personalized test papers for users in a targeted manner to meet their diverse learning needs by determining the test paper composition intention and portrait information of the user, determining multiple test paper composition types that match the user based on the portrait information, and determining the recommended weights of each test question under each test paper composition type based on the test paper composition intention. Since the user's learning level, knowledge mastery, learning preferences and other factors, as well as the user's assessment objectives and content scope are fully considered, the generated test papers are more in line with the user's actual situation, and can better help users to check for omissions, consolidate knowledge, and improve their abilities, thereby improving the efficiency and quality of test paper composition, enhancing the user's learning experience, realizing personalized intelligent test paper composition, effectively improving students' learning efficiency and learning effects, and avoiding the problems of traditional test paper composition methods that are not targeted and cannot meet the personalized needs of users, as well as the problems of low efficiency and difficult to ensure quality of manual test paper composition.

[0069] Based on the above embodiment, multiple test type combinations matching the user are determined based on the profile information, including: Based on the portrait information, multiple test paper types that match the user are determined, as well as the difficulty ratio of knowledge points under each test paper type.

[0070] Specifically, the difficulty ratio of knowledge points refers to the proportion of knowledge points of different difficulty levels under a specific test paper type. For example, in the "basic consolidation type" test paper type, the proportion of simple difficulty knowledge points may be higher, while the proportion of difficult difficulty knowledge points may be lower; in the "ability improvement type" test paper type, the proportion of difficult difficulty knowledge points may be higher, while the proportion of simple difficulty knowledge points may be lower.

[0071] Taking into account the differences in learning levels and knowledge mastery among different users, and the different test paper types having different assessment objectives and content focuses, the difficulty ratio of knowledge points must also be considered when setting up the test paper to ensure that the generated test paper can better meet the personalized needs of users and achieve the corresponding assessment objectives. The difficulty ratio of knowledge points can then be dynamically adjusted based on the user's portrait information and the test paper type to improve the pertinence and effectiveness of the test paper, and avoid the generated test paper being too difficult or too low, which will result in users being unable to effectively evaluate their own learning situation or unable to achieve the corresponding learning goals.

[0072] As an optional embodiment, a default knowledge point difficulty ratio can be set for each test type according to predefined rules. Then, the default knowledge point difficulty ratio can be adjusted based on the user's historical scores and weak knowledge points. For example, if the user's weak knowledge points are concentrated in a certain difficulty level, the knowledge point ratio of that difficulty level can be appropriately increased. Finally, the knowledge point difficulty ratio can be further optimized based on the user's feedback information, and the knowledge point difficulty ratio under each test type can be finally determined. Among them, the above-mentioned predefined rules can be based on the characteristics of the test type itself to set the default knowledge point difficulty ratio. For example, the default knowledge point difficulty ratio of the "basic consolidation type" test is "easy: medium: difficult = 70%: 20%: 10%", the default knowledge point difficulty ratio of the "ability improvement type" test is "easy: medium: difficult = 30%: 40%: 30%", and the default knowledge point difficulty ratio of the "pre-exam sprint type" test is "easy: medium: difficult = 20%: 50%: 30%".

[0073] As another optional embodiment, a machine learning model, such as a regression model or a classification model, can also be used to predict the optimal knowledge point difficulty ratio of the user under the test type based on the user's portrait information and the test type; the training data of the model can include the user's historical learning data, the learning data of other users, and the knowledge point difficulty ratio set by experts, etc., to ultimately determine the knowledge point difficulty ratio under each test type.

[0074] In addition, considering that the user's learning situation may change over time and that there may be certain deviations in the user's feedback information, after determining the difficulty ratio of the knowledge points, the difficulty ratio of the knowledge points can also be regularly evaluated and adjusted. For example, by analyzing the user's test-taking situation, test scores, and feedback information, it can be determined whether the current knowledge point difficulty ratio is reasonable, and corresponding adjustments can be made based on the evaluation results.

[0075] Based on any of the above embodiments, Figure 3 This is a flow chart of the implementation of step 132 in the test paper composition method provided by the present invention. Figure 3 As shown, step 132 determines the weight of each question under each test type based on the test paper composition intention, including: Step 1321: Determine the first recommendation weight of each question in each test paper type based on the difficulty ratio of the knowledge points in each test paper type. Step 1322: Based on the test paper setting intention, determine the second recommendation weight of each test question under each test paper setting type; Step 1323: The first recommendation weight and the second recommendation weight of each test question are integrated to obtain the recommendation weight of each test question.

[0076] Specifically, the first recommendation weight refers to the degree of recommendation of the test question determined based on the knowledge point difficulty ratio under a specific test paper type, which is used to characterize the degree of match between the difficulty of the test question and the difficulty distribution expected by the test paper type. Since the knowledge point difficulty ratio reflects the proportion that knowledge points of different difficulty levels should occupy under the test paper type, the higher the first recommendation weight, the more the difficulty of the test question meets the overall difficulty requirements of the test paper type, and the first recommendation weight determined based on the knowledge point difficulty ratio can ensure that the overall difficulty structure of the test paper meets the preset requirements. For example, for the "basic consolidation type" test paper type, the first recommendation weight of test questions with lower difficulty may be relatively high.

[0077] As an optional embodiment, the first recommendation weight of the test question can be calculated based on the difficulty level of the test question and the expected proportion of the difficulty level under the test paper type. For example, if the difficulty level of a test question is "easy" and the expected proportion of the "easy" difficulty level under the test paper type is 70%, the first recommendation weight of the test question can be set to 0.7 to determine the first recommendation weight of each test question.

[0078] As another optional embodiment, a fuzzy matching method can be used to fuzzily match the difficulty level of the test question with the expected difficulty distribution for the test paper type, and the first recommendation weight of the test question can be determined based on the degree of match. For example, a Gaussian function can be used to describe the expected difficulty distribution for the test paper type, and then the similarity between the difficulty level of the test question and the Gaussian function can be calculated. This similarity is the first recommendation weight of the test question, and the first recommendation weight of each test question is determined.

[0079] In addition, the second recommendation weight refers to the degree of recommendation of the test questions determined based on the user's test-taking intention under a specific test-taking type, which is used to characterize the degree of match between the test questions and the test-taking requirements explicitly specified by the user. Since the test-taking intention includes the user's test-taking requirements in multiple different dimensions, such as knowledge points, question types, local test conditions, etc., the higher the second recommendation weight, the more the test question meets the user's personalized test-taking requirements, and the second recommendation weight determined based on the test-taking intention can ensure that the content structure of the test paper meets the user's specific needs. For example, if the user specifies that the "Kinetic Energy Theorem" should be examined in detail, then the second recommendation weight of the test questions that examine knowledge points related to the "Kinetic Energy Theorem" will be relatively high.

[0080] As an optional embodiment, the second recommendation weight of the test question can be calculated by matching the knowledge points, question types, local test conditions and other attributes tested by the test question with the test paper composition requirements specified by the user; for example, the keyword matching method can be used to calculate the number of user-specified keywords contained in the test question description. The greater the number, the higher the second recommendation weight of the test question, and the second recommendation weight of each test question is determined.

[0081] As another optional embodiment, a deep learning model can also be used to perform semantic analysis on the test questions and the user's test paper composition requirements, and calculate the semantic similarity between them. The semantic similarity is the second recommendation weight of the test question, and the second recommendation weight of each test question is determined.

[0082] Consider that if the test paper is generated based solely on the first recommended weight of each question, it may ignore the user's personalized needs, resulting in a uniform test paper that fails to meet the user's specific learning goals, and thus leads to a poor user experience. For example, even in the "Basic Consolidation" test type, different users may want to focus on different knowledge points.

[0083] Furthermore, considering that if the test paper is compiled based solely on the second recommended weight of each question, the overall difficulty structure of the test paper may be ignored, resulting in an unreasonable difficulty distribution of the generated test paper, which cannot effectively assess the user's learning level and thus makes it difficult to ensure the quality of the test paper. For example, even if the user specifies that they want to focus on a certain knowledge point, it is necessary to ensure that the test paper has a certain proportion of questions of different difficulty levels.

[0084] Based on this, the embodiment of the present invention integrates the first recommendation weight and the second recommendation weight of each test question to obtain the final recommendation weight of each test question, thereby taking into account the overall difficulty structure of the test paper and the personalized needs of the user, ensuring the quality of the test paper while meeting the user's specific learning goals, and avoiding problems caused by only considering one aspect while ignoring the other. For example, for a test question that tests the "Kinetic Energy Theorem", if its difficulty level is "medium", and the user specifies that the "Kinetic Energy Theorem" should be the focus, then its final recommendation weight will be calculated by combining its first recommendation weight (reflecting the degree of "medium" difficulty level) and the second recommendation weight (reflecting the degree of testing the "Kinetic Energy Theorem").

[0085] As an optional embodiment, a linear weighting method can be used to perform a weighted summation of the first recommendation weight and the second recommendation weight of each test question to obtain the final recommendation weight of each test question; for example, the final recommendation weight = α × first recommendation weight + (1-α) × second recommendation weight, where α is an adjustable parameter used to control the relative importance of the first recommendation weight and the second recommendation weight to obtain the recommendation weight of each test question. For example, if the user has high requirements for the overall difficulty of the test paper, but does not place special emphasis on the examination of specific knowledge points, a higher value (such as 0.7) can be assigned to α to place more emphasis on the first recommendation weight and ensure that the overall difficulty structure of the test paper meets the preset requirements. If the user explicitly specifies that certain knowledge points should be examined, but has no special requirements for the overall difficulty of the test paper, a lower value (such as 0.3) can be assigned to α to place more emphasis on the second recommendation weight and ensure that the content structure of the test paper meets the specific needs of the user.

[0086] Based on any of the above embodiments, determining the second recommendation weight of each question under each test paper type based on the test paper setting intention includes: Determine the initial recommended weight of each question based on the knowledge point type of each question under each test paper type; Based on the test paper setting intention, the initial weight of each test question is adjusted to obtain the second recommended weight of each test question.

[0087] Specifically, the knowledge point type of each question refers to the attributes of the knowledge point tested by the question. For example, the knowledge point type of each question can include core knowledge points, frequently tested points, points prone to error, weak points, and newly tested points. Each question in the question bank can be pre-labeled with a corresponding type tag, and the corresponding knowledge point type can be obtained based on the type tag of each question.

[0088] The initial recommendation weight refers to the degree of recommendation determined based on the knowledge point type of the test question, without considering the user's personalized test-taking intention. For example, for a test-taking type of "special practice on local high-frequency test points", high-frequency test points can be given a higher initial recommendation weight.

[0089] The initial recommendation weight of each test question can be determined based on preset rules. For example, for core knowledge points, the initial recommendation weight is set to 0.8; for high-frequency test points, the initial recommendation weight is set to 0.7; for easy-to-make mistakes, the initial recommendation weight is set to 0.6; for weak points, the initial recommendation weight is set to 0.5; for newly added test points, the initial recommendation weight is set to 0.4. In addition, an expert system can also be used to set different initial recommendation weights for test questions of different knowledge point types based on factors such as the importance of the knowledge point, test frequency, and student mastery, to determine the initial recommendation weight of each test question. For example, the expert system can set different initial recommendation weights for test questions of different knowledge point types based on the analysis results of previous years' real questions.

[0090] In addition, considering that the initial recommendation weight of each test question only takes into account the knowledge point type of the test question itself, but does not take into account the user's personalized test paper composition intention, if the test paper is composed only based on the initial recommendation weight of each test question, it may cause the generated test paper to fail to meet the user's specific needs. For example, the user may want to focus on a specific knowledge point, or want to avoid testing a knowledge point that has already been mastered, which will result in the test paper being less targeted and unable to effectively help users improve their learning effects.

[0091] Based on this, the embodiment of the present invention adjusts the initial weight of each question based on the test-setting intent, so that the resulting second recommended weight not only reflects the importance of the knowledge point type of the question itself, but also meets the user's personalized test-setting intent. For example, if the user specifies that they want to focus on the "Kinetic Energy Theorem," the recommended weight of questions that test knowledge points related to the "Kinetic Energy Theorem" can be increased; if the user specifies that they want to avoid testing a knowledge point that they have already mastered, the recommended weight of questions that test knowledge points related to that knowledge point can be reduced.

[0092] As an optional embodiment, keywords can be first extracted according to the test paper intention specified by the user; then, the similarity between the description text of each test question and these keywords can be calculated; then, the initial recommendation weight of each test question can be adjusted according to the similarity. For example, the second recommendation weight can be determined using the following formula: second recommendation weight = initial recommendation weight × (1 + similarity); finally, the second recommendation weights of all test questions are normalized so that their values ​​are between 0 and 1, thereby obtaining the second recommendation weight of each test question.

[0093] Based on any of the above embodiments, the recommended weight of each question under each test paper type is determined based on the test paper setting intention, including: Based on the test-taking intention, determine the user's test-taking needs in multiple dimensions; Based on the test paper requirements under each dimension, determine the recommended weight of each test question under each test paper type.

[0094] Specifically, test paper requirements across different dimensions refer to users' specific requirements for the test paper in terms of knowledge points, difficulty, question types, and exam context. For example, a user may want a test paper that focuses on the "Kinetic Energy Theorem" knowledge point, with a "medium" difficulty level, a mix of "multiple choice" and "essay questions," and includes "local high-frequency test points." Because users may have preferences and needs across different dimensions, test paper creation requires comprehensive consideration of user requirements across all dimensions to generate a test paper that truly meets their expectations. Otherwise, considering only one or a few dimensions may result in a test paper that fails to meet the user's overall requirements.

[0095] In order to further improve the accuracy and personalization of test paper generation, better meet the needs of users in different dimensions, and avoid the deviation between the generated test papers and the user's expectations, the embodiment of the present invention first determines the user's test paper generation needs in multiple different dimensions based on the test paper generation intention, so as to more comprehensively understand the user's test paper generation goals, more accurately grasp the user's test paper generation preferences, more effectively improve the quality and efficiency of test paper generation, and realize truly personalized intelligent test paper generation.

[0096] As an optional embodiment, natural language processing technology can be used to analyze the user's test-taking intentions, extract keywords and key information, and then map these keywords and key information to different dimensions to determine the user's test-taking needs in multiple different dimensions. For example, named entity recognition technology can be used to identify entities such as knowledge points, difficulty, and question types in the user's intentions, and classify them into corresponding dimensions to determine the user's test-taking needs in multiple different dimensions.

[0097] After determining the test paper requirements across different dimensions, a comprehensive score can be calculated for each question based on the user's needs across these dimensions. This comprehensive score is then used as the recommended weight for the question, yielding the recommended weights for each question across each test paper type. For example, a weighted summation approach can be used to weight the scores across different dimensions for each question. The weights can be adjusted based on the user's emphasis on each dimension; if the user places particular importance on a particular dimension, the weight for that dimension can be increased.

[0098] Among them, considering that the test questions in the question bank usually need to be marked in advance, the marked information includes the knowledge points tested by the test questions, difficulty levels, question types, etc., the test questions in the question bank can be pre-processed before determining the recommendation weight of each test question. For example, knowledge graph technology can be used to perform more fine-grained division and association of the knowledge points tested by the test questions, so as to more accurately calculate the degree of match between the test questions and user needs.

[0099] Specifically, the knowledge graph can be used to organize knowledge points into a network, establishing hierarchical relationships between knowledge points (for example, "force" → "Newton's Laws of Motion" → "Newton's Second Law") and associations (for example, "kinetic energy theorem" and "law of conservation of energy"). This allows users to automatically expand to related knowledge points when specifying a knowledge point, thereby expanding the scope of topic selection and ensuring the knowledge point coverage of the exam paper. Furthermore, each question in the question bank is associated with a knowledge point in the knowledge graph, allowing for one question to correspond to multiple knowledge points, or one knowledge point to correspond to multiple questions. The weight of the association can be determined based on the degree to which the question examines the knowledge point. Finally, the knowledge graph can be used for knowledge reasoning. For example, if a user wishes to test their understanding of the "kinetic energy theorem," knowledge reasoning can be used to find questions that require the "kinetic energy theorem" to solve. The knowledge graph can support multi-dimensional knowledge point matching, for example, matching not only the name of the knowledge point but also the concept, attributes, and application scenarios of the knowledge point. The knowledge graph can also be updated dynamically. For example, it can automatically adjust the association and weight between knowledge points based on the user's test performance, test scores and other data, so that the knowledge graph is more in line with the user's actual situation.

[0100] Based on any of the above embodiments, the test paper type carries a priority tag; the test paper is compiled based on the recommended weight of each test question under each test paper type to obtain a personalized test paper for the user, including: Based on the recommended weights of each question under each test paper type, the test paper is compiled to obtain personalized test papers corresponding to each test paper type; Based on the priority labels of each test paper type, the personalized test papers corresponding to each test paper type are prioritized.

[0101] Specifically, the priority label of each test paper type refers to an identifier used to distinguish the importance or recommendation order of different test paper types, which is used to indicate which test paper type is more in line with the user's current needs or learning goals among multiple optional test paper types. Among them, the priority label of each test paper type can be determined based on the user's portrait information. For example, based on the user's portrait information, the user can be divided into three categories: excellent, average, and poor. For users of the excellent type, the priority label of their test paper type is "special practice papers for weak points > standard papers > advanced papers, special practice papers for easy-to-make mistakes > special practice papers for frequently tested points"; for users of the average type, the priority label of their test paper type is "special practice papers for weak points > special practice papers for easy-to-make mistakes > standard papers, special practice papers for frequently tested points > advanced papers"; for users of the poor type, the priority label of their test paper type is "special practice papers for weak points > special practice papers for easy-to-make mistakes > standard papers, special practice papers for frequently tested points > advanced papers".

[0102] In addition, the user's historical learning data, test-taking records, and feedback information can be combined to use a machine learning model to predict the user's preference for different test-taking types, and the preference level can be used as a priority label for the test-taking type. This embodiment of the present invention does not specifically limit this.

[0103] Considering that different types of test papers focus on different assessment objectives and contents, and different users have different needs for different test paper types, if the priority of the test paper types is not distinguished, and the test papers generated by all test paper types are treated equally, it may cause users to be at a loss when choosing test papers, or mistakenly select test papers that do not meet their needs, resulting in poor user experience and failure to effectively improve learning outcomes.

[0104] Based on this, the embodiment of the present invention first compiles the test papers based on the recommended weights of each test question under each test paper type, obtains the test papers corresponding to each test paper type, and then prioritizes the test papers corresponding to each test paper type based on the priority labels of each test paper type, such as sorting the test papers corresponding to each test paper type in order of priority from high to low, thereby more effectively guiding users to select the test papers that best meet their needs, improving the utilization rate and learning effect of the test papers, better meeting the personalized learning needs of users, and avoiding users from feeling confused or making wrong choices when selecting test papers. Among them, the test papers corresponding to each test paper type refer to the test question set generated according to the recommended weight of the test paper type, which can be a complete test paper or a question package containing multiple test questions. The embodiment of the present invention does not make specific limitations on this.

[0105] For example, for a user with excellent academic performance, if the three types of test papers, "Weak Points Training Paper", "Standard-reaching Paper" and "Excellence-enhancing Paper" are generated at the same time, then when they are displayed to the user, the "Weak Points Training Paper" will be placed first because this type is more in line with the current learning needs of the excellent user.

[0106] It is understood that after the personalized test papers corresponding to each set of test paper types are prioritized, personalized test papers can be recommended to users based on the priority ranking, or personalized test papers can be displayed in the test paper list in order of priority, thereby guiding users to prioritize personalized test papers that best meet their needs, thereby improving test paper utilization and learning outcomes. For example, the personalized test paper with the highest priority can be displayed prominently on the homepage, or personalized test papers can be arranged in order of priority in the test paper list.

[0107] Based on any of the above embodiments, a test paper is generated based on the recommended weights of each test question under each test paper type to obtain a personalized test paper for the user, and then the following steps are further included: Receive tuning instructions from users; Based on the tuning information carried by the tuning instructions, the recommended weight of each test question under each test paper type is optimized; The test paper is compiled based on the optimized recommended weights of each test question to obtain an optimized personalized test paper.

[0108] Specifically, the user's tuning instruction refers to the instruction for adjusting the test paper composition strategy issued when the user is dissatisfied with the generated test paper. It is usually generated after the user previews or uses the generated test paper. For example, the user may feel that the test paper is too difficult, or the knowledge points are not fully covered, or the question types do not meet their needs, thereby issuing a tuning instruction. Among them, the generation method of the tuning instruction may include: ① The user manually adjusts the difficulty coefficient, knowledge point weight, question type ratio and other parameters of the test paper. ② The user replaces or deletes individual questions in the test paper. ③ The user gives natural language instructions, such as "too difficult", "incomplete knowledge point coverage", "question type does not meet the requirements", etc., and the test paper composition strategy is automatically adjusted according to the user's natural language instructions. The above is an example of the generation method of the tuning instruction. The embodiment of the present invention does not specifically limit the generation method of the tuning instruction.

[0109] After receiving the tuning instruction, it indicates that the user is not satisfied with the currently generated test paper and hopes to make adjustments based on their feedback to generate a test paper that better meets their needs. In this case, based on the tuning information carried by the tuning instruction, the recommendation weight of each test question under each test paper type is optimized. For example, if the user feels that the difficulty of the test paper is too high, the recommendation weight of the more difficult test questions can be lowered, and the recommendation weight of the less difficult test questions can be increased. Among them, the tuning information refers to the information provided by the user in the tuning instruction to guide the adjustment of the test paper strategy, which may include parameter values ​​manually adjusted by the user, question information replaced or deleted by the user, and the user's evaluation of the test paper.

[0110] As an optional embodiment, the tuning instructions can be parsed first to extract the tuning information. Then, based on the tuning information, the recommended weights of each question in each test type can be adjusted. For example, if the user manually adjusts the difficulty coefficient, the recommended weights of questions of each difficulty level can be adjusted accordingly based on the adjustment ratio of the difficulty coefficient. Finally, the adjusted recommended weights are normalized to a value between 0 and 1, thereby optimizing the recommended weights of each question in each test type.

[0111] After optimizing the recommended weights for each question in each test type, the test is re-generated based on the optimized recommended weights to generate a new test paper, resulting in an optimized test paper. This optimized test paper better meets the user's needs and can better meet the user's learning goals, thereby improving the user's learning experience and learning results.

[0112] It is understandable that if the optimized test paper still fails to fully meet the user's needs, the user can issue tuning instructions again and provide more detailed feedback information to generate tuning instructions again, and continue to optimize the recommended weights of each test question under each test paper type, so that the optimized test paper is more and more in line with the user's actual situation, and finally generate a personalized test paper that fully meets the user's expectations.

[0113] Based on any of the above embodiments, the intention of the test preparation instruction is analyzed to obtain the user's test preparation intention, including: Based on the portrait information, multiple candidate test composition intentions are determined; Based on the examination test instruction, the user's examination test intention is determined from multiple candidate examination test intentions.

[0114] Specifically, a candidate test-setting intention refers to the objectives and content range of the test paper that a user may wish to generate, inferred based on the user's profile information. It is an alternative to the user's actual test-setting intention. Since the profile information may include the user's grade, subject, history score, weak knowledge points, frequently tested knowledge points, study time, current textbook version, textbook volume, etc., the user's learning stage and learning content can be analyzed based on the profile information, and the user's learning goals and learning needs can be inferred, thereby generating multiple candidate test-setting intentions. For example, if the user is a third-grade elementary school student studying the "addition and subtraction" unit, it can be inferred that the user may need candidate test-setting intentions such as "basic addition and subtraction exercises" and "addition and subtraction word problem exercises." If the user is a second-grade junior high school student with a midterm exam approaching, it can be inferred that the user may need candidate test-setting intentions such as "midterm review paper" and "weak knowledge point reinforcement paper," resulting in multiple candidate test-setting intentions.

[0115] After identifying multiple candidate test-taking intents, the system can use a large model to perform semantic analysis on the test-taking instructions, extracting the keywords and key information contained in the instructions. These keywords and key information are then matched with the candidate test-taking intents, and the similarity between each candidate test-taking intent and the test-taking instruction is calculated. The candidate with the highest similarity is selected as the user's true test-taking intent. For example, if the user enters the instruction "Midterm exam is coming up, please generate a review paper for me," the system can calculate the similarity between this instruction and candidate test-taking intents such as "Midterm review paper" and "Weak knowledge point reinforcement paper," ultimately determining that "Midterm review paper" is the user's true test-taking intent.

[0116] Furthermore, to improve the accuracy of intent recognition, a multi-round dialogue approach can be used to interact with the user, guiding them to select from multiple candidate test-taking intents. For example, a user could be presented with candidate test-taking intents such as "Midterm Review" and "Weak Points Reinforcement" and asked, "Which type of test paper would you prefer?" The final test-taking intent is determined based on the user's selection.

[0117] Figure 4 This is the second flow chart of the test paper composition method provided by the present invention. Figure 4As shown, the method includes: first, based on the user's login information on a learning device (such as a learning machine), determining the user's profile information. This profile information includes the user's grade, the current textbook version, and the textbook volume. Based on the user's profile information, multiple candidate test-setting intentions are determined, each corresponding to at least one test paper, such as "Help me find a moving point on a number axis," "Help me find the shortest path of an axially symmetric problem," or "Help me find a customized midterm test for the X-edition X-grade (first half) midterm." Next, the user selects a final test-setting intention from the multiple candidate test-setting intentions, and the corresponding test paper is generated based on the test-setting intention.

[0118] Figure 5 This is the third flow chart of the test paper composition method provided by the present invention, as shown in FIG. Figure 5 As shown, the method includes: displaying multiple function entrances (such as new curriculum standards, customized papers, easy-to-make mistakes, weak questions, frequently tested questions), etc. on the display interface of a learning device (such as a learning machine). If any function entrance is selected by the user, it indicates that the user's intention to create a test paper is to select questions within the scope represented by the function entrance. For example, if the user selects "new curriculum standards", it indicates that the user's intention to create a test paper is to generate a test paper that meets the requirements of the new curriculum standards. After the user selects any function entrance, multiple candidate test paper creation intentions under the corresponding function entrance are displayed in combination with the user's portrait information, so that the user can select the final test paper creation intention from the multiple candidate test paper creation intentions. After the test paper creation intention is determined, the test paper creation intention is input into the large model, and the large model outputs the corresponding test paper.

[0119] For example, for the "New Curriculum Standards", the corresponding candidate test-setting intentions may include "Find new situational literacy questions for Unit N", where Unit N is selected from the "Units with new course titles". For "Customized Test Papers", the corresponding candidate test-setting intentions may include "Find customized test papers for Unit N", and Unit N may be displayed according to the month of the semester the user is in, trying to match the user's learning progress as much as possible, for example, N is displayed as 5 in January, 1 in February, 2 in March, and 3 in April, etc. For "Easy-to-make mistakes", the corresponding candidate test-setting intentions may include "Find easy-to-make mistakes questions for Unit N for special practice", and Unit N may also be displayed according to the month of the semester the user is in, trying to match the user's learning progress as much as possible. For "Weak Questions", the corresponding candidate test-setting intentions may include "Help me find the weak points of Unit N for special practice". For "Frequently Tested Questions", the corresponding candidate test-setting intentions may include "Help me find frequently tested questions for Unit N for special practice".

[0120] Figure 6 This is the fourth flow chart of the test paper composition method provided by the present invention, as shown in FIG. Figure 6As shown, the method includes: obtaining user profile information and test-setting intentions. The test-setting intentions can be determined by: ① Voice dialogue: The user interacts with the system through voice to express their requirements for the test paper, such as "I want to do a set of exercises on the kinetic energy theorem"; voice recognition technology is used to convert the user's voice into text, and the user's intention is further analyzed to extract keywords and key information. ② Text: The user directly enters text in the input box to express their requirements for the test paper, such as "Generate a high school physics final review paper"; natural language processing technology is used to analyze the user's intention and extract keywords and key information. ③ Preset Intentions: Common test-taking intentions are pre-set, such as "Final Review," "Strengthening Weak Points," and "Practice on Frequently Tested Points." Users can directly select these pre-set intentions to quickly generate test papers. These intentions are then fed into the big model, which analyzes the user's profile, the difficulty of each knowledge point in the question bank (easy, medium, and difficult), the type of each knowledge point in the question bank (including frequently tested points, points prone to error, and weak points), and the local exam landscape (including the comprehensive exam situation in the user's area and relevant knowledge points covered in important local midterm and final exams) to generate the appropriate test paper. If the user is dissatisfied with the output, they can re-enter the test-taking intention using the three methods mentioned above, such as through multiple rounds of dialogue, until a satisfactory test paper is obtained.

[0121] The following describes the roll assembly device provided by the present invention. The roll assembly device described below and the roll assembly method described above can be referenced to each other.

[0122] Based on any of the above embodiments, Figure 7 This is a schematic diagram of the structure of the paper assembly device provided by the present invention. Figure 7 As shown, the device includes: Receiving unit 710, for receiving a test preparation instruction from a user; An analysis unit 720 is used to perform an intention analysis on the test preparation instruction to obtain the user's test preparation intention; The customization unit 730 is used to customize the user's personalized test paper based on the test paper composition intention and the user's profile information.

[0123] Based on any of the above embodiments, the user's personalized test paper is customized based on the test paper composition intention and the user's profile information, including: Based on the profile information, multiple test type combinations matching the user are determined; Based on the test paper setting intention, determine the recommended weight of each question under each test paper type; The test paper is compiled based on the recommended weight of each question under each test paper type to obtain the user's personalized test paper.

[0124] Based on any of the above embodiments, multiple test type combinations matching the user are determined based on the profile information, including: Based on the portrait information, multiple test paper types that match the user are determined, as well as the difficulty ratio of knowledge points under each test paper type.

[0125] Based on any of the above embodiments, the weight of each question under each test type is determined based on the test paper composition intention, including: Based on the difficulty ratio of knowledge points in each test paper type, determine the first recommendation weight of each test question in each test paper type; Based on the test paper setting intention, determine the second recommendation weight of each question under each test paper setting type; The first recommendation weight and the second recommendation weight of each test question are integrated to obtain the recommendation weight of each test question.

[0126] Based on any of the above embodiments, determining the second recommendation weight of each question under each test paper type based on the test paper setting intention includes: Determine the initial recommended weight of each question based on the knowledge point type of each question under each test paper type; Based on the test paper setting intention, the initial weight of each test question is adjusted to obtain the second recommended weight of each test question.

[0127] Based on any of the above embodiments, the test paper type carries a priority tag; the test paper is compiled based on the recommended weight of each test question under each test paper type to obtain a personalized test paper for the user, including: Based on the recommended weights of each question under each test paper type, the test paper is compiled to obtain personalized test papers corresponding to each test paper type; Based on the priority labels of each test paper type, the personalized test papers corresponding to each test paper type are prioritized.

[0128] Based on any of the above embodiments, a test paper is generated based on the recommended weights of each test question under each test paper type to obtain a personalized test paper for the user, and then the following steps are further included: Receive tuning instructions from users; Based on the tuning information carried by the tuning instructions, the recommended weight of each test question under each test paper type is optimized; The test paper is compiled based on the optimized recommended weights of each test question to obtain an optimized personalized test paper.

[0129] Based on any of the above embodiments, performing intent analysis on the test preparation instruction to obtain the user's test preparation intention includes: Based on the portrait information, multiple candidate test composition intentions are determined; Based on the examination test instruction, the user's examination test intention is determined from multiple candidate examination test intentions.

[0130] Figure 8 Schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 8As shown, the electronic device may include: a processor 810, a communications interface 820, a memory 830, and a communication bus 840. The processor 810, the communications interface 820, and the memory 830 communicate with each other via the communications bus 840. The processor 810 may invoke logic instructions in the memory 830 to execute a test paper generation method, which includes: receiving a test paper generation instruction from a user; performing an intent analysis on the test paper generation instruction to obtain the user's test paper generation intention; and customizing a personalized test paper for the user based on the test paper generation intention and the user's profile information.

[0131] Furthermore, the logic instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0132] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the test paper setting method provided by the above methods, which includes: receiving the user's test paper setting instructions; performing intention analysis on the test paper setting instructions to obtain the user's test paper setting intention; and customizing the user's personalized test paper based on the test paper setting intention and the user's portrait information.

[0133] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it is implemented to execute the test paper assembling method provided by the above-mentioned methods, the method comprising: receiving a user's test paper assembling instruction; performing intention analysis on the test paper assembling instruction to obtain the user's test paper assembling intention; and customizing the user's personalized test paper based on the test paper assembling intention and the user's portrait information.

[0134] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0135] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A test paper composition method, characterized in that: include: Receive the user's test preparation instruction; Performing intention analysis on the test-taking instruction to obtain the test-taking intention of the user; Based on the test paper composition intention and the user's portrait information, a personalized test paper for the user is customized.

2. The test paper composition method according to claim 1, characterized in that: The step of customizing the user's personalized test paper based on the test paper composition intention and the user's portrait information includes: Based on the portrait information, determining multiple test type combinations that match the user; Based on the test paper composition intention, determining the recommended weight of each test question under each test paper composition type; The test paper is compiled based on the recommended weight of each test question under each test paper type to obtain a personalized test paper for the user.

3. The test paper composition method according to claim 2, characterized in that: The determining of multiple test type combinations matching the user based on the portrait information includes: Based on the portrait information, multiple test paper types that match the user are determined, as well as the difficulty ratio of knowledge points under each test paper type.

4. The test paper composition method according to claim 3, characterized in that: Determining the weight of each question under each test paper type based on the test paper composition intention includes: Based on the difficulty ratio of knowledge points in each test paper type, determine the first recommendation weight of each test question in each test paper type; Based on the test paper setting intention, determining the second recommendation weight of each test question under each test paper setting type; The first recommendation weight and the second recommendation weight of each test question are integrated to obtain the recommendation weight of each test question.

5. The test paper composition method according to claim 4, characterized in that: The determining, based on the test paper setting intention, of the second recommendation weight of each test question under each test paper setting type includes: Determine the initial recommended weight of each question based on the knowledge point type of each question under each test paper type; Based on the test paper composition intention, the initial weight of each test question is adjusted to obtain a second recommended weight for each test question.

6. The test paper assembling method according to any one of claims 1 to 5, characterized in that: The test paper type carries a priority tag; and the test paper is compiled based on the recommended weight of each test question under each test paper type to obtain the user's personalized test paper, including: Based on the recommended weights of each question under each test paper type, the test paper is compiled to obtain personalized test papers corresponding to each test paper type; Based on the priority labels of each test paper type, the personalized test papers corresponding to each test paper type are prioritized.

7. The test paper assembling method according to any one of claims 1 to 5, characterized in that: The test paper is composed based on the recommended weight of each test question under each test paper type to obtain the user's personalized test paper, and then further includes: receiving a tuning instruction from the user; Based on the tuning information carried by the tuning instruction, optimizing the recommended weight of each test question under each test paper type; The test paper is compiled based on the optimized recommended weights of each test question to obtain an optimized personalized test paper.

8. The test paper assembling method according to any one of claims 1 to 5, characterized in that: The performing intention analysis on the test taking instruction to obtain the user's test taking intention includes: Based on the portrait information, determining multiple candidate test composition intentions; Based on the roll formation instruction, the roll formation intention of the user is determined from the multiple candidate roll formation intentions.

9. A roll-forming device, characterized in that: include: A receiving unit, configured to receive a test paper generating instruction from a user; An analysis unit, configured to perform an intention analysis on the test-taking instruction to obtain the test-taking intention of the user; The customization unit is used to customize the user's personalized test paper based on the test paper composition intention and the user's portrait information.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the test paper assembling method according to any one of claims 1 to 8 is implemented.

11. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the test paper assembling method according to any one of claims 1 to 8 is implemented.

12. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the test paper assembling method according to any one of claims 1 to 8 is implemented.

Images