Online question bank intelligent test question recommendation method and system

Abstract

The invention relates to the technical field of online education, in particular to an intelligent test question recommendation method and system for an online question bank, and the method comprises the steps: obtaining a test request which carries a knowledge unit; generating a question information pool; a plurality of question information is stored in the database, and the question information comprises questions, question examination points, the number of examination points, knowledge units to which the questions belong, historical average question answering duration and historical accuracy; calculating the difficulty coefficient of each piece of called question information; generating a question set; receiving difficulty rating of a user on the knowledge unit; generating a question selection framework, wherein the preset question framework comprises a plurality of question examination points, the question number of each examination point, the question sequence number of each question and the difficulty coefficient range of each question sequence number; and selecting question information of the adaptive difficulty coefficient in the question set according to the question selection framework to generate a test question group. The method has the effect of generating the test question group according with the actual learning level of the student according to the historical learning record of the student and the difficulty rating of the knowledge unit.

Description

Technical Field

[0001] The present invention relates to the technical field of online education, and in particular to an online question bank intelligent test question recommendation method and system. Background Art

[0002] In the field of education, with the vigorous development of Internet technology, various online education platforms and software have provided students with unprecedented learning resources and convenient learning methods. As an important tool for evaluating students' learning effects, the importance of online question banks has become increasingly significant. However, online question banks in the existing technology still have many deficiencies in terms of test question recommendation, especially when dealing with common test question types such as multiple-choice questions, true-or-false questions, and fill-in-the-blank questions. These problems are particularly prominent, affecting students' learning experience and the accuracy of effect evaluation.

[0003] Specifically, existing online question bank systems usually adopt a relatively simple test question push strategy, such as randomly selecting questions to form a test question group, or pushing questions according to preset questions. This push method ignores the actual learning situation and individual differences of students, and it is difficult to accurately reflect the learning and absorption results of students. For students with weak foundations, test questions that are too difficult may make them feel frustrated, thereby affecting their learning enthusiasm; while for students with good foundations, test questions that are too easy may not be able to effectively test their learning outcomes and lack challenges.

[0004] In addition, the existing online question bank system lacks intelligence and personalization in terms of question recommendation. Although some systems provide the function of adjusting the difficulty of questions based on the students' historical answering situation, this method is often too simple and only relies on a single indicator such as the accuracy rate, while ignoring other key factors that may affect the difficulty of the questions. Summary of the invention

[0005] In order to generate a test question set that meets the students' actual learning level based on their historical learning records and difficulty ratings of knowledge units, so as to accurately detect the students' learning situation, the present application provides an online question bank intelligent test question recommendation method and system.

[0006] The above-mentioned invention objective of the present application is achieved through the following technical solutions:

[0007] An online question bank intelligent test question recommendation method, comprising:

[0008] Obtaining a test request, wherein the test request carries a knowledge unit;

[0009] Based on the knowledge unit, the question information under the knowledge unit is retrieved in the database to generate a question information pool; the database stores a plurality of question information, and the question information includes the question, the question test point, the number of test points, the knowledge unit to which the question belongs, the historical average question-solving time and the historical correct rate;

[0010] Calculate the difficulty coefficient of each question information retrieved, where the difficulty coefficient of the question information is calculated based on the number of test points, the historical average test time and the historical accuracy rate;

[0011] Generate a question set, the question set including a plurality of question elements, each question element including question information and a corresponding difficulty coefficient;

[0012] receiving a user's difficulty rating of the knowledge unit;

[0013] Generate a question selection framework according to the difficulty rating, the total difficulty range and a preset question framework, wherein the preset question framework includes a plurality of question test points, the number of questions at each test point, a question number of each question, and a difficulty coefficient range of each question number;

[0014] The test question group is generated by selecting question information of appropriate difficulty coefficient from the question set according to the question selection framework.

[0015] In a preferred example, the present application can be further configured as follows: the difficulty coefficient of the question information is calculated based on the number of test points, the historical average test time and the historical accuracy rate, and is specifically calculated using the following formula:

[0016]

[0017] Among them, a is the historical accuracy rate, b is the historical average time for taking questions, c is the difficulty coefficient, d is the number of test points, τ is the accuracy influence coefficient, μ is the accuracy weight factor, v is the logarithmic adjustment factor, δ is the logarithmic influence coefficient, γ is the difficulty scaling factor, θ is the time and test point adjustment coefficient, and ρ is the test point difficulty influence coefficient.

[0018] In a preferred example, the present application can be further configured as follows: the value range of τ is 0.5~3; the value range of μ is 0.5~3; the value range of v is 10~200; the value range of δ is 0.8 or 1.5; the value range of γ is 0.8~1.5; the value range of θ is 0.1~1; the value range of ρ is 0.1~0.6.

[0019] In a preferred example, the present application can be further configured as follows: the value of τ is 1; the value of μ is 1; the value of v is 10; the value of δ is 1; the value of γ is 1; the value of θ is 0.2; and the value of ρ is 0.2.

[0020] In a preferred example, the present application can be further configured as follows: the topic selection framework includes multiple topic test points, the number of topics for each test point, the number of topics for each test point, and the topic selection difficulty coefficient range of each topic sequence number. The topic selection difficulty coefficient range of each topic sequence number is obtained by the following method:

[0021] A value range is calculated based on the difficulty rating in the overall difficulty level;

[0022] According to the value range, a sub-range is intercepted in the difficulty coefficient range of each question number to generate the selection difficulty coefficient range of each question number.

[0023] In a preferred example, the present application can be further configured as follows: selecting the topic information of the appropriate difficulty coefficient from the topic set according to the topic selection framework to generate a test question group, including:

[0024] If there are multiple question information in the same range of question difficulty coefficient for a question number, the question information with the largest number of test points among the multiple question information in the same range of question difficulty coefficient is selected as the question information of the question number.

[0025] The second object of the invention is achieved by the following technical solutions:

[0026] An online question bank intelligent test question recommendation system, comprising:

[0027] A request generation module, used for obtaining a test request, wherein the test request carries a knowledge unit;

[0028] An information pool generation module is used to retrieve the question information under the knowledge unit in the database based on the knowledge unit to generate a question information pool; the database stores a plurality of question information, and the question information includes the question, the question test point, the number of test points, the knowledge unit to which the question belongs, the historical average question-solving time and the historical correct rate;

[0029] A difficulty coefficient calculation module is used to calculate the difficulty coefficient of each question information retrieved, where the difficulty coefficient of the question information is calculated based on the number of test points, the historical average question-taking time and the historical correct rate;

[0030] A question set generation module, used to generate a question set, wherein the question set includes a plurality of question elements, each of which includes question information and a corresponding difficulty coefficient;

[0031] A rating module, used to receive the user's difficulty rating of the knowledge unit;

[0032] A question selection framework generation module, used to generate a question selection framework according to the difficulty rating, the total difficulty range and a preset question framework, wherein the preset question framework includes a plurality of question test points, the number of questions at each test point, a question serial number of each question, and a difficulty coefficient range of each question serial number;

[0033] The question set generation module is used to generate a test question set by selecting question information of an appropriate difficulty coefficient from a question set according to a question selection framework.

[0034] In summary, the present application includes at least one of the following beneficial technical effects:

[0035] 1. The system can generate test questions that meet the students' actual learning level for common test question types such as multiple-choice questions, true-or-false questions, and fill-in-the-blank questions based on the students' historical learning records and the difficulty rating of the knowledge units. It can accurately detect the students' learning and absorption results, and can also provide test questions of appropriate difficulty to stimulate students' learning interest and enthusiasm, thereby improving learning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a flowchart of an online question bank intelligent question recommendation method in one embodiment of the present application;

[0037] Figure 2 This is a flowchart of a specific subdivision of step S6 in the online question bank intelligent question recommendation method in one embodiment of the present application;

[0038] Figure 3 This is a module connection diagram of an online question bank intelligent test question recommendation system in one embodiment of the present application. DETAILED DESCRIPTION

[0039] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0040] It should be noted that the terms "first", "second", etc. in the present invention are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure.

[0041] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0042] Reference Figure 1 and Figure 2 , the online question bank intelligent question recommendation method includes:

[0043] S1. Obtain a test request, which carries a knowledge unit.

[0044] The system or platform can automatically or manually obtain the test request of the learner in a variety of ways. The request clearly identifies the information of the knowledge unit to be tested, which is the knowledge unit that the learner is studying. The specific ways of generating test requests include but are not limited to the following: automatic triggering, manual request, timed reminder, and achievement unlocking.

[0045] Among them, the automatic trigger type is that after the student completes all learning tasks of a certain knowledge unit on the learning platform (such as watching videos, completing reading), the system automatically identifies the learning progress and immediately generates a test request containing the information of the knowledge unit. For example, if the student completes all the learning content of the "Basic Mathematics - Fractional Operations" module, the system automatically generates a test request containing "Basic Mathematics - Fractional Operations" as a test knowledge point.

[0046] Manual request type is that the student actively clicks the "Test" button on the learning interface, and the system generates a corresponding test request based on the knowledge unit information of the current learning page. For example, after the student has finished learning the "Programming Basics - Loop Structure" page, he clicks the "Start Test" button, and the system generates a test request for "Programming Basics - Loop Structure" based on the page information.

[0047] For the timed reminder type, the system sends a test reminder to the student according to the preset learning plan or time interval. The student confirms and then takes the test. At this time, the test request also contains the corresponding knowledge unit. For example, the system is set to remind the student to take the test of the "English Conversation - Daily Communication" module at 4 pm every day. After the student receives the reminder and confirms, the system generates a test request.

[0048] The achievement unlocking method is that when students achieve certain learning achievements (such as completing a certain number of exercises, obtaining a certain score, etc.), the system automatically unlocks the test permission for the next knowledge unit and generates a test request. For example, if a student completes 100 exercises in the "Principles of Physics - Basic Mechanics" module and the average score reaches 80 points, the system automatically unlocks the test for "Principles of Physics - Newton's Laws of Motion" and generates a test request.

[0049] S2. Based on the knowledge unit, the question information under the knowledge unit is retrieved from the database to generate a question information pool.

[0050] The database stores information about a plurality of questions, including the question, the test points of the question, the number of test points, the knowledge unit to which the question belongs, the historical average time for answering the question, and the historical accuracy rate.

[0051] After receiving a test request carrying a specific knowledge unit, the system enters this step, the core of which is to accurately retrieve the question information associated with the knowledge unit from the pre-built database to build a question information pool for the test.

[0052] Each piece of question information contains detailed metadata, including the question, the test points of the question, the number of test points, the knowledge unit to which the question belongs, the average time for answering the question in history, and the historical accuracy rate.

[0053] Among them, the title refers to the specific content of the test question. The test point of the question refers to the knowledge point or skill point involved in the question, usually one or more. The number of test points refers to the total number of test points covered in the question, which helps to evaluate the complexity and coverage of the question. The knowledge unit to which the question belongs refers to the knowledge unit that clearly identifies the question to which the question belongs, so that the system can accurately match it according to the test request. The historical average time for answering the question refers to the average time for students to answer the question in the past, which can be used as a reference for evaluating the difficulty of the question and the learning status of the students. It is understandable that when the student answers the question, the time from the end of the student's answering the previous question to the end of the current question is recorded as the student's time for answering the question. The historical accuracy rate refers to the accuracy rate of students' answers to the question in the past, reflecting the discrimination of the question and the general mastery of the students. According to the knowledge unit information carried in the test request, the system performs a query operation in the database, filters out all the question information belonging to the knowledge unit, and combines them into a question set.

[0054] S3. Calculate the difficulty coefficient of each question information retrieved. The difficulty coefficient of the question information is calculated based on the number of test points, the historical average question-taking time, and the historical accuracy rate.

[0055] Among them, the difficulty coefficient of each question information is calculated based on the number of test points, the historical average test time and the historical accuracy rate, and is calculated using the following formula:

[0056]

[0057] Among them, a is the historical accuracy rate, b is the historical average time for taking questions, c is the difficulty coefficient, d is the number of test points, τ is the accuracy influence coefficient, μ is the accuracy weight factor, v is the logarithmic adjustment factor, δ is the logarithmic influence coefficient, γ is the difficulty scaling factor, θ is the time and test point adjustment coefficient, and ρ is the test point difficulty influence coefficient.

[0058] Specifically, τ is the accuracy influence coefficient, which adjusts the nonlinear influence of the accuracy a on the difficulty c; μ is the accuracy weight factor, which further adjusts the weight of the accuracy a when calculating the difficulty c; ν is the logarithmic adjustment factor, which controls the speed of change of the internal value of the logarithmic function, thereby affecting the logarithmic adjustment of the question time b and the number of test points d on the difficulty c; δ is the logarithmic influence coefficient, which adjusts the degree of influence of the logarithmic function on the difficulty c; η is the difficulty scaling factor, which is used to adjust the overall size of the difficulty; θ is the time and test point adjustment coefficient, which is used to adjust the influence of the number of test points d and the question time b on the internal value of the logarithmic function. is the test point difficulty influence coefficient, d ρ It directly reflects the positive impact of the number of test points d on the difficulty of the questions c. Used to adjust the degree of this effect. When d increases, the difficulty c will also increase.

[0059] The value range of τ is 0.5~3; the value range of μ is 0.5~3; the value range of v is 10~200; the value range of δ is 0.8 or 1.5; the value range of γ is 0.8~1.5; the value range of θ is 0.1~1; the value range of ρ is 0.1~0.6.

[0060] Based on the formula constructed above, various coefficients are trained through a large number of question data samples, and relevant characteristic variables are pre-extracted from each question data sample, such as the number of test points d, historical accuracy a, historical average time for taking questions b, and pre-calibrated difficulty coefficient c, etc., and trained through machine learning models, such as nonlinear regression models or neural networks. In this embodiment, neural networks are used for training.

[0061] The neural network is trained through a sample set of labeled question data to obtain a formula model, thereby obtaining the value of each coefficient within the value range of each coefficient.

[0062] In a specific embodiment, the values ​​of the coefficients are: τ is 1; μ is 1; v is 10; δ is 1; γ is 1; θ is 0.2; and ρ is 0.2.

[0063] After building the question set, the core of this step is to use metadata such as the number of test points, historical average test time, and historical accuracy in the question information to calculate the difficulty coefficient of each question through a specific mathematical model. By comprehensively considering multiple dimensions such as the number of test points, historical average test time, and historical accuracy, the system can more accurately assess the difficulty of each question, providing a scientific basis for intelligent test composition and personalized testing. By calculating the difficulty coefficient, test questions can be managed more efficiently, making it easier to classify, update, and maintain the difficulty of questions, thereby improving the timeliness and accuracy of test content.

[0064] S4. Generate a question set, the question set includes multiple question elements, each question element includes question information and a corresponding difficulty coefficient.

[0065] S5. Receive the user's difficulty rating for the knowledge unit.

[0066] Specifically, the system pre-sets multiple difficulty ratings A, B, C, D, and E, among which difficulty rating A corresponds to the 1-20% interval, difficulty rating B corresponds to the 21-40% interval, difficulty rating C corresponds to the 41%-60% interval, difficulty rating D corresponds to the 61-80% interval, and difficulty rating E corresponds to the 81-100% interval.

[0067] S6. Generate a question selection framework based on the difficulty rating, total difficulty range and preset question framework.

[0068] Among them, the preset test question framework includes multiple test points, the number of questions in each test point, the question number of each question, and the difficulty coefficient range of each question number.

[0069] The topic selection framework includes multiple test points, the number of questions in each test point, the number of questions in each test point and the question serial number, as well as the range of the difficulty coefficient of each question serial number.

[0070] The topic selection framework is as follows:

[0071] Test point 1: Linear equations of one variable. Number of questions: 3. Question numbers: 1.1, 1.2, 1.3. Difficulty range of questions: 1.1: 2-3 (basic difficulty); 1.2: 4-5 (medium difficulty); 1.3: 6-7 (higher difficulty).

[0072] Test point 2: Basic properties of triangles. Number of questions: 4. Question numbers: 2.1, 2.2, 2.3, 2.4. Difficulty range of questions: 2.1: 3-4 (basic difficulty); 2.2: 5-6 (medium difficulty); 2.3: 7-8 (higher difficulty); 2.4: 9-10 (high difficulty).

[0073] Test point 3: addition, subtraction, multiplication and division of fractions. Number of questions: 2. Question numbers: 3.1, 3.2. Difficulty range: 3.1: 2-4 (basic to medium difficulty); 3.2: 6-8 (medium to high difficulty).

[0074] Among them, the range of difficulty coefficients of each question number is obtained by the following method:

[0075] S61. Calculate a value range in the total difficulty level according to the difficulty rating.

[0076] The interval is 1-100%, and the value range is obtained by the corresponding interval of the above difficulty rating.

[0077] S62. According to the value range, a sub-range is extracted from the difficulty coefficient range of each question number to generate a selection difficulty coefficient range of each question number.

[0078] The accuracy rate a ranges from [0.01, 1]; the historical average test time b ranges from [1, 600] seconds; the number of test points d is a positive integer ranging from [1, 10].

[0079] After calculation and statistics, the difficulty coefficient range is basically distributed between 0 and 10. For example, when a=0.1, b=10, d=1, the value of the difficulty coefficient C is about 6.67.

[0080] Therefore, the difficulty coefficient range can be set from 0 to 10.

[0081] The range of the difficulty coefficient of the question number is intercepted from the original difficulty coefficient range through the value interval. For example, in the difficulty coefficient range of 6-8 (medium to high difficulty), there are five sub-ranges of 6-6.40, 6.41-6.80, 6.81-7.20, 7.21-7.60, and 7.61-8; if the value interval corresponding to the student's difficulty rating is larger, the smaller sub-range is intercepted. For example, if the corresponding value interval is 61-80%, the range of the difficulty coefficient of the question is 6.41-6.80; if the corresponding value interval is 1-20%, the range of the difficulty coefficient of the question is 7.61-8. Furthermore, by accurately matching the difficulty coefficient range according to the student's difficulty rating, the difficulty selection of the question is further refined to avoid losing the challenge because the question is too simple, or causing frustration because the question is too difficult.

[0082] S7. Selecting question information of an appropriate difficulty coefficient from the question set according to the question selection framework to generate a test question group.

[0083] Specifically, if there are multiple question information in the same range of question difficulty coefficients for a question number, the question information with the largest number of test points among the multiple question information in the same range of question difficulty coefficients is selected as the question information of the question number. Furthermore, the question coverage can be improved. By selecting the question information with the largest number of test points, it can be ensured that students can be exposed to more different test points when practicing. This helps students to master the knowledge points more comprehensively and improve the breadth and depth of learning. Questions with a large number of test points usually contain more information and details, which helps students to exercise their thinking ability and problem-solving skills in the process of answering. This comprehensive question has a positive effect on improving students' learning effects and their ability to deal with complex problems. Improve teaching efficiency; and help ensure that students get the maximum learning benefits in a limited time and promote knowledge integration.

[0084] This application also provides an online question bank intelligent test question recommendation system, refer to Figure 3 ,include:

[0085] The request generation module is used to obtain a test request, and the test request carries a knowledge unit.

[0086] The information pool generation module is used to retrieve the question information under the knowledge unit in the database based on the knowledge unit to generate a question information pool; the database stores multiple question information, and the question information includes the question, the question test points, the number of test points, the knowledge unit to which the question belongs, the historical average question-solving time and the historical accuracy rate.

[0087] The difficulty coefficient calculation module is used to calculate the difficulty coefficient of each retrieved question information. The difficulty coefficient of the question information is calculated based on the number of test points, the historical average question-taking time and the historical accuracy rate.

[0088] The question set generation module is used to generate a question set, which includes multiple question elements, each of which includes question information and a corresponding difficulty coefficient.

[0089] The rating module is used to receive the difficulty rating of the knowledge unit from the user.

[0090] The question selection framework generation module is used to generate a question selection framework according to the difficulty rating, the total difficulty range and the preset question framework. The preset question framework includes multiple question test points, the number of questions in each test point, the question serial number of each question, and the difficulty coefficient range of each question serial number.

[0091] The question set generation module is used to generate a test question set by selecting question information of an appropriate difficulty coefficient from a question set according to a question selection framework.

[0092] The specific definition of the online question bank intelligent question recommendation system can be found in the above definition of the online question bank intelligent question recommendation method, which will not be repeated here. Each step of the above online question bank intelligent question recommendation method can be implemented in whole or in part by software, hardware or a combination thereof.

[0093] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0094] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for programmable processors and can be implemented using high-level procedural and / or object-oriented programming languages ​​and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or means (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0095] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0096] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0097] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0098] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. An online question bank intelligent question recommendation method, characterized in that: include: Obtaining a test request, wherein the test request carries a knowledge unit; Based on the knowledge unit, the topic information under the knowledge unit is retrieved in the database to generate a topic information pool; The database stores a plurality of question information, the question information including the question, the question test points, the number of test points, the knowledge unit to which the question belongs, the historical average time for answering the question and the historical accuracy rate; Calculate the difficulty coefficient of each question information retrieved, where the difficulty coefficient of the question information is calculated based on the number of test points, the historical average test time and the historical accuracy rate; Generate a question set, the question set including a plurality of question elements, each question element including question information and a corresponding difficulty coefficient; receiving a user's difficulty rating of a knowledge unit; Generate a question selection framework according to the difficulty rating, the total difficulty range and a preset question framework, wherein the preset question framework includes a plurality of question test points, the number of questions at each test point, a question number of each question, and a difficulty coefficient range of each question number; The test question group is generated by selecting question information of appropriate difficulty coefficient from the question set according to the question selection framework.

2. The online question bank intelligent question recommendation method according to claim 1, characterized in that: The difficulty coefficient of the question information is calculated based on the number of test points, the historical average test time and the historical accuracy rate, and is specifically calculated using the following formula: Among them, a is the historical accuracy rate, b is the historical average time for taking questions, c is the difficulty coefficient, d is the number of test points, τ is the accuracy influence coefficient, μ is the accuracy weight factor, v is the logarithmic adjustment factor, δ is the logarithmic influence coefficient, γ is the difficulty scaling factor, θ is the time and test point adjustment coefficient, and ρ is the test point difficulty influence coefficient.

3. The online question bank intelligent question recommendation method according to claim 2, characterized in that: The value range of τ is 0.5~3; the value range of μ is 0.5~3; the value range of v is 10~200; the value range of δ is 0.8 or 1.5; the value range of γ is 0.8~1.5; the value range of θ is 0.1~1; the value range of ρ is 0.1~0.

6.

4. The online question bank intelligent question recommendation method according to claim 2, characterized in that: The value of τ is 1; the value of μ is 1; the value of v is 10; the value of δ is 1; the value of γ is 1; the value of θ is 0.2; and the value of ρ is 0.

2.

5. The online question bank intelligent question recommendation method according to claim 1, characterized in that: The topic selection framework includes multiple topic test points, the number of topics in each test point, the topic number of each test point, and the topic difficulty coefficient range of each topic number. The topic difficulty coefficient range of each topic number is obtained in the following way: A value range is calculated based on the difficulty rating in the overall difficulty level; According to the value range, a sub-range is intercepted in the difficulty coefficient range of each question number to generate the selection difficulty coefficient range of each question number.

6. The online question bank intelligent question recommendation method according to claim 5, characterized in that: According to the topic selection framework, the topic information of the appropriate difficulty coefficient is selected from the topic set to generate a test question group, including: If there are multiple question information in the same range of question difficulty coefficient for a question number, the question information with the largest number of test points among the multiple question information in the same range of question difficulty coefficient is selected as the question information of the question number.

7. An online question bank intelligent question recommendation system based on any online question bank intelligent question recommendation method in claims 1-6, characterized in that: include: A request generation module, used for obtaining a test request, wherein the test request carries a knowledge unit; An information pool generation module is used to retrieve the question information under the knowledge unit in the database based on the knowledge unit to generate a question information pool; the database stores a plurality of question information, and the question information includes the question, the question test point, the number of test points, the knowledge unit to which the question belongs, the historical average question-solving time and the historical correct rate; A difficulty coefficient calculation module is used to calculate the difficulty coefficient of each question information retrieved, where the difficulty coefficient of the question information is calculated based on the number of test points, the historical average question-taking time and the historical correct rate; A question set generation module, used to generate a question set, wherein the question set includes a plurality of question elements, each of which includes question information and a corresponding difficulty coefficient; A rating module, used to receive the user's difficulty rating of the knowledge unit; A question selection framework generation module, used to generate a question selection framework according to the difficulty rating, the total difficulty range and a preset question framework, wherein the preset question framework includes a plurality of question test points, the number of questions at each test point, a question serial number of each question, and a difficulty coefficient range of each question serial number; The question set generation module is used to generate a test question set by selecting question information of an appropriate difficulty coefficient from a question set according to a question selection framework.

Images