Intelligent question setting system based on domain knowledge graph

Abstract

The invention belongs to the technical field of intelligent question setting, particularly relates to an intelligent question setting system based on a domain knowledge graph, and provides the following scheme for solving the problem that the learning effect is reduced due to the fact that the difficulty of test questions is inconvenient to intelligently adjust according to the question answering condition of students and the generated test questions are inconvenient to evaluate in the prior art. The system comprises a data acquisition module; the knowledge graph construction module is connected with a domain screening module, the domain screening module is connected with a data cleaning module, the data cleaning module is connected with the data acquisition module, and the knowledge graph module is connected with a database module; the extraction module is connected with a question generation module, in the use process, the test question difficulty can be intelligently adjusted according to the question answering situation of students, the generated test questions can be evaluated conveniently, and therefore the learning effect can be effectively guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of intelligent question-setting technology, and in particular to an intelligent question-setting system based on a domain knowledge graph. Background Art

[0002] At present, AI-enabled education has become an important feature of higher education reform. Reshaping the higher education ecosystem through AI technology and developing intelligent education have become new issues in higher education at present and even in the future. Different from traditional education, intelligent education uses intelligent technologies represented by AI to accelerate the transformation of professional talent training models and realize the organic combination of large-scale education and personalized training. Knowledge graphs are an indispensable basic resource for AI, with rich semantic expression capabilities, unified description forms, and support for automatic acquisition of knowledge. Various industries are actively exploring knowledge graphs based on vertical fields to enhance the intelligence level of education, finance and other industries through knowledge empowerment; the academic community is committed to studying the automation methods of various knowledge graph construction and application. Compared with general knowledge graphs, domain knowledge graphs are oriented to specific fields and built based on industry data. They have strict and rich data models and have higher requirements for the depth and accuracy of domain knowledge. The question-setting system is a system that can automatically generate test questions. It uses computer technology and artificial intelligence algorithms to help educators, test organizers, etc. create test questions quickly and efficiently.

[0003] In the existing technology, it is not convenient to intelligently adjust the difficulty of test questions according to students' answers, and it is not convenient to evaluate the generated test questions, which leads to reduced learning effect. For this reason, we propose an intelligent question-setting system based on domain knowledge graph to solve the above problems. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings in the prior art that it is inconvenient to intelligently adjust the difficulty of test questions according to students' answers, and it is inconvenient to evaluate the generated test questions, thereby reducing the learning effect, and propose an intelligent question-setting system based on domain knowledge graph.

[0005] The intelligent question-setting system based on domain knowledge graph provided in this application adopts the following technical solutions:

[0006] An intelligent question-setting system based on domain knowledge graph, comprising:

[0007] Data acquisition module;

[0008] A knowledge graph construction module, wherein the knowledge graph construction module is connected to a field screening module, wherein the field screening module is connected to a data cleaning module, wherein the data cleaning module is connected to a data acquisition module, and wherein the knowledge graph module is connected to a database module;

[0009] An extraction module, wherein the extraction module is connected to a topic generation module, wherein the topic generation module is connected to an evaluation module, wherein the extraction module is connected to a database module, wherein the database module is connected to a database management module;

[0010] A login module, wherein the login module is connected to a selection module, the selection module is connected to an intelligent recommendation module, the intelligent recommendation module is connected to a matching module, the matching module is connected to a question bank management module, the question bank management module is connected to a question database module, and the question bank management module is connected to a question generation module;

[0011] An analysis module, wherein the analysis module is connected to a recording module, the recording module is connected to a practice module, the practice module is connected to a matching module, and the analysis module is connected to an intelligent recommendation module.

[0012] Furthermore, the topic generation module includes a knowledge point selection unit, a template design unit and a generation unit, the knowledge point selection unit is connected to the template design unit, and the template design unit is connected to the generation unit.

[0013] Furthermore, the analysis module includes an error question analysis unit, a progress analysis unit and a difficulty analysis unit, the error question analysis unit is connected to the progress analysis unit, and the progress analysis unit is connected to the difficulty analysis unit.

[0014] Furthermore, the knowledge graph construction module is used to organize knowledge points, concepts, formulas, instance data and their relationships into a graph to form a knowledge graph, and the database module is used to store the generated knowledge graph data.

[0015] Furthermore, the question bank management module is used to establish a rich question bank, supporting classification and labeling storage of questions, and the question database module is used to store the generated questions.

[0016] Furthermore, the practice module is used for users to answer questions, the recording module is used to record the answering data, and the analysis module is used to analyze correct and incorrect questions, answering progress and answering difficulty.

[0017] Furthermore, the extraction module is used to select specific knowledge points from the knowledge graph in the database module as the basis for setting questions, and the question generation module is used to generate corresponding questions based on the extracted knowledge points.

[0018] Furthermore, the data collection module is used to collect knowledge points, concepts, formulas, and instance data; the data cleaning module is used to clean and remove duplicates from the collected data; and the field screening module is used to screen out single field knowledge.

[0019] Furthermore, the evaluation module includes a semantic detection unit, a comparison unit and a judgment unit, wherein the semantic detection unit is connected to the comparison unit, and the comparison unit is connected to the judgment unit.

[0020] Furthermore, the question bank management module includes a question type classification unit, a difficulty classification unit and a labeling storage unit, the question type classification unit is connected to the difficulty classification unit, and the difficulty classification unit is connected to the labeling storage unit.

[0021] In summary, the present application includes at least one of the following beneficial technical effects:

[0022] 1. This solution is based on the setting of analysis module and intelligent recommendation module. The analysis module can analyze the wrong questions, answering progress and answering difficulty. The intelligent recommendation module can intelligently adjust the difficulty of the test questions according to the analysis results and recommend corresponding test questions.

[0023] 2. This solution sets up an evaluation module, which can use a scoring algorithm to evaluate the matching degree between the generated questions and the standard answers, as well as detect the semantics of the questions to ensure the accuracy of the questions.

[0024] During use, the present invention can facilitate intelligent adjustment of the difficulty of test questions according to the students' answers, and facilitate evaluation of the generated test questions, thereby effectively ensuring the learning effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a structural diagram of an intelligent question-setting system based on domain knowledge graph proposed by the present invention;

[0026] Figure 2 This is a structural block diagram of a question bank management module of an intelligent question setting system based on domain knowledge graph proposed by the present invention;

[0027] Figure 3 This is a structural block diagram of a question generation module of an intelligent question setting system based on domain knowledge graph proposed by the present invention;

[0028] Figure 4 This is a structural block diagram of an evaluation module of an intelligent question-setting system based on domain knowledge graph proposed by the present invention;

[0029] Figure 5 This is a structural block diagram of the analysis module of an intelligent question-setting system based on domain knowledge graph proposed in the present invention. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0031] Embodiment 1

[0032] Reference Figure 1-Figure 5 , an intelligent question-setting system based on domain knowledge graph, comprising:

[0033] Data acquisition module;

[0034] A knowledge graph construction module, the knowledge graph construction module is connected to a domain screening module, the domain screening module is connected to a data cleaning module, the data cleaning module is connected to a data acquisition module, and the knowledge graph module is connected to a database module;

[0035] An extraction module, the extraction module is connected to a topic generation module, the topic generation module is connected to an evaluation module, the extraction module is connected to a database module, and the database module is connected to a database management module;

[0036] A login module, the login module is connected to a selection module, the selection module is connected to an intelligent recommendation module, the intelligent recommendation module is connected to a matching module, the matching module is connected to a question bank management module, the question bank management module is connected to a question database module, and the question bank management module is connected to a question generation module;

[0037] Analysis module, the analysis module is connected to the recording module, the recording module is connected to the practice module, the practice module is connected to the matching module, the analysis module is connected to the intelligent recommendation module, the data collection module is used to collect knowledge points, concepts, formulas, and instance data, the data cleaning module is used to clean and remove duplicates from the collected data, the field screening module is used to screen out single field knowledge, the knowledge graph construction module is used to organize knowledge points, concepts, formulas, instance data and their relationships into a graph to form a knowledge graph, the database module is used to store the generated knowledge graph data, and the intelligent recommendation module adopts a machine learning algorithm, which includes a linear regression algorithm and a logistic regression algorithm;

[0038] 1. Linear regression algorithm: Linear regression is a prediction model whose goal is to predict the relationship between a dependent variable and one or more independent variables by fitting an optimal straight line (or hyperplane). The formula of linear regression can be expressed as:

[0039] y=β0+β1x1+β2x2+...+∈

[0040] Where y is the dependent variable, x1, x2, ... are independent variables, β0, β1, β2, ... are regression coefficients, ∈ is the error term, and the purpose of linear regression is to estimate these parameters by minimizing the mean squared error (MSE);

[0041] 2. Logistic regression algorithm: Logistic regression is used to solve binary classification problems. Its output range is between 0 and 1. It is usually used to predict the probability of an event. The formula of logistic regression can be expressed as:

[0042]

[0043] where Y is a binary dependent variable (0 or 1), x1, x2, ..., x n are the independent variables, β0,β1,β2,...,β n are the logistic regression coefficients. Logistic regression estimates these parameters by maximizing the likelihood function.

[0044] Reference Figure 1 and Figure 2 The question bank management module includes a question type classification unit, a difficulty classification unit and a labeling storage unit. The question type classification unit is connected to the difficulty classification unit, and the difficulty classification unit is connected to the labeling storage unit. The question bank management module is used to establish a rich question bank and supports the classification and labeling storage of questions. The question database module is used to store the generated questions.

[0045] Reference Figure 1 and Figure 3 The question generation module includes a knowledge point selection unit, a template design unit and a generation unit. The knowledge point selection unit is connected to the template design unit, and the template design unit is connected to the generation unit. The extraction module is used to select specific knowledge points from the knowledge graph in the database module as the basis for setting questions. The question generation module is used to generate corresponding questions based on the extracted knowledge points.

[0046] Reference Figure 4 ,The evaluation module includes a semantic detection unit, a control unit and a judgment unit. The semantic detection unit is connected with the control unit, and the control unit is connected with the judgment unit. The evaluation module adopts a scoring algorithm. The scoring algorithm formula mainly includes the calculation of the difficulty coefficient and the discrimination coefficient.

[0047] The difficulty coefficient reflects the difficulty of the test questions and is usually calculated by the following formula:

[0048] 1. For objective test questions, the difficulty P calculation formula is: Where k is the number of people who answered the question correctly, and N is the total number of people who took the test;

[0049] 2. For subjective questions, the difficulty P calculation formula is: Where X is the average score of the test, and M is the full score of the test;

[0050] 3. There is also a calculation formula applicable to subjective and objective test questions: PH and PL are the difficulty values ​​of the test questions for high-scoring and low-scoring candidates respectively;

[0051] The discrimination coefficient reflects the ability of the test item to distinguish subjects of different levels and is calculated by the following formula:

[0052] 1. Sort the scores from high to low. The first 50% of the candidates are the high score group, and the last 50% are the low score group;

[0053] 2. The calculation formula of the discrimination coefficient Dr is: Ah is the average score of the high-scoring group, Al is the average score of the low-scoring group, and T is the full score.

[0054] Reference Figure 1 and Figure 5 The analysis module includes an error analysis unit, a progress analysis unit and a difficulty analysis unit. The error analysis unit is connected to the progress analysis unit, and the progress analysis unit is connected to the difficulty analysis unit. The practice module is used for users to answer questions. The recording module is used to record the answering data. The analysis module is used to analyze wrong questions, answering progress and answering difficulty.

[0055] The implementation principle of this embodiment is as follows: when in use, the data collection module collects knowledge points, concepts, formulas, and instance data; the data cleaning module can clean and remove duplicates from the collected data; the field screening module can screen out single field knowledge; and then the knowledge graph construction module can organize the knowledge points, concepts, formulas, instance data and their relationships into a graph to form a knowledge graph; the database module can store the generated knowledge graph data; the extraction module selects specific knowledge points from the knowledge graph in the database module as the basis for setting questions; the question generation module is used to generate corresponding questions based on the extracted knowledge points, and the generated questions are evaluated by the evaluation module. Evaluation, to ensure the quality of test questions, the question bank management module can classify and label the generated test questions for storage, the question database module can store the generated questions, the user logs in through the login module, selects learning questions through the selection module, or makes intelligent recommendations based on the intelligent recommendation module, matches the test questions in the question database module through the matching module, and then answers the questions through the practice module, the recording module can record the answer data, the analysis module can analyze the wrong questions, the answer progress and the answer difficulty, and transmit the analysis results to the intelligent recommendation module, the intelligent recommendation module intelligently adjusts the difficulty of the test questions according to the machine learning algorithm, and recommends the corresponding test questions.

[0056] Embodiment 2

[0057] The difference between this embodiment and the first embodiment is that the question generation module is connected to the intelligent question selection module, which can generate questions that meet the teaching progress according to the teaching syllabus and knowledge points, including multiple-choice questions, fill-in-the-blank questions, short-answer questions, etc., which comprehensively cover the knowledge points.

[0058] Embodiment 3

[0059] The difference between this embodiment and the first embodiment is that the question generation module is connected to the intelligent test paper module, which supports multiple test paper modes and can customize the test paper structure according to needs, including the number of questions, difficulty distribution, etc., to meet diversified test paper needs.

[0060] Embodiment 4

[0061] The difference between this embodiment and the first embodiment is that the knowledge graph construction module is connected to the real-time update module, which can continuously update the education resource library, introduce new knowledge points and question types in a timely manner, and maintain the advancement and practicality of the system.

[0062] Embodiment 5

[0063] The difference between this embodiment and the first embodiment is that the question generation module is connected to the feedback optimization module, and the feedback optimization module can collect user feedback in actual applications and continuously optimize and improve the system based on the feedback.

[0064] Experimental example

[0065] By comparing the intelligent question-setting solutions proposed in Examples 1 to 5 with conventional intelligent question-setting solutions, the experimental data are as follows:

[0066]

[0067]

[0068] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent question-setting system based on domain knowledge graph, characterized by: include: Data acquisition module; A knowledge graph construction module, wherein the knowledge graph construction module is connected to a field screening module, wherein the field screening module is connected to a data cleaning module, wherein the data cleaning module is connected to a data acquisition module, and wherein the knowledge graph module is connected to a database module; An extraction module, wherein the extraction module is connected to a topic generation module, wherein the topic generation module is connected to an evaluation module, wherein the extraction module is connected to a database module, wherein the database module is connected to a database management module; A login module, wherein the login module is connected to a selection module, the selection module is connected to an intelligent recommendation module, the intelligent recommendation module is connected to a matching module, the matching module is connected to a question bank management module, the question bank management module is connected to a question database module, and the question bank management module is connected to a question generation module; An analysis module, wherein the analysis module is connected to a recording module, the recording module is connected to a practice module, the practice module is connected to a matching module, and the analysis module is connected to an intelligent recommendation module.

2. According to claim 1, an intelligent question-setting system based on domain knowledge graph is characterized in that: The question bank management module includes a question type classification unit, a difficulty classification unit and a labeling storage unit. The question type classification unit is connected to the difficulty classification unit, and the difficulty classification unit is connected to the labeling storage unit.

3. According to claim 2, an intelligent question-setting system based on domain knowledge graph is characterized in that: The topic generation module includes a knowledge point selection unit, a template design unit and a generation unit. The knowledge point selection unit is connected to the template design unit, and the template design unit is connected to the generation unit.

4. The intelligent question-setting system based on domain knowledge graph according to claim 3 is characterized by: The evaluation module comprises a semantic detection unit, a comparison unit and a judgment unit, wherein the semantic detection unit is connected to the comparison unit, and the comparison unit is connected to the judgment unit.

5. The intelligent question-setting system based on domain knowledge graph according to claim 4 is characterized by: The analysis module includes an error question analysis unit, a progress analysis unit and a difficulty analysis unit. The error question analysis unit is connected to the progress analysis unit, and the progress analysis unit is connected to the difficulty analysis unit.

6. The intelligent question-setting system based on domain knowledge graph according to claim 5, characterized in that: The data collection module is used to collect knowledge points, concepts, formulas, and instance data; the data cleaning module is used to clean and remove duplicates from the collected data; and the field screening module is used to screen out single field knowledge.

7. The intelligent question-setting system based on domain knowledge graph according to claim 6, characterized in that: The knowledge graph construction module is used to organize knowledge points, concepts, formulas, instance data and their relationships into a graph to form a knowledge graph, and the database module is used to store the generated knowledge graph data.

8. The intelligent question-setting system based on domain knowledge graph according to claim 7, characterized in that: The extraction module is used to select specific knowledge points from the knowledge graph in the database module as the basis for setting questions, and the question generation module is used to generate corresponding questions based on the extracted knowledge points.

9. The intelligent question-setting system based on domain knowledge graph according to claim 8, characterized in that: The question bank management module is used to establish a rich question bank, supporting classification and labeling storage of questions, and the question database module is used to store generated questions.

10. The intelligent question-setting system based on domain knowledge graph according to claim 9, characterized in that: The practice module is used for users to answer questions, the recording module is used to record answering data, and the analysis module is used to analyze correct and incorrect questions, answering progress and answering difficulty.

Images