Course teaching optimization method and system based on deep learning

Abstract

The invention discloses a course teaching optimization method and system based on deep learning, and belongs to the technical field of data learning processing, and the method comprises the steps: 1, obtaining the learning data of a preset student, the learning data comprising learning behaviors and academic records of a preset subject; based on the academic record of the preset subject, analyzing the mastering degree of each knowledge point by the preset student; 2, dividing the knowledge points according to low-difficulty knowledge points, medium-difficulty knowledge points and high-difficulty knowledge points according to the mastering condition, and obtaining preset learning duration of students for the medium-difficulty knowledge points and the high-difficulty knowledge points; 2, dividing the knowledge points according to the mastering conditions, and obtaining the learning duration of the students for the medium-difficulty knowledge points and the high-difficulty knowledge points; by means of accurate evaluation, personalized teaching, learning behavior optimization, dynamic adjustment and the like, the teaching effect and the learning efficiency of students can be effectively improved, and powerful support is provided for improving the teaching quality.

Description

Technical Field

[0001] The present invention relates to the technical field of data learning and processing, and particularly to a course teaching optimization method and system based on deep learning. Background Art

[0002] With the rapid development of educational informatization, traditional teaching models are facing numerous challenges. Nowadays, with the rapid development of artificial intelligence, deep learning and neural network technologies have become the core driving forces for educational innovation. The AI neural network deep learning integrated teaching experiment and training system has emerged, providing cutting-edge teaching methods for universities and vocational colleges.

[0003] Among them, the application document with the technical application number CN201811107772.X provides an optimization method for the teaching of advanced mathematics surface courses. This method uses a surface demonstration teaching aid, which includes two support members. The support member includes a base, a column, a fixing rod, a fixing column, a support ring, and a support plate. The column is fixedly installed on the top of the base, and the upper ends of the columns of the two support members are connected by a fixing rod. One fixing column is fixedly connected to one side of the upper end of the column. A support ring is provided on the outer circle of the fixing column, and a support plate is provided at the bottom of the support ring. The support ring is concentric with the fixing column, and an annular groove is formed between the support ring and the fixing column. Expansion modules are provided in the two opposite annular grooves. This method can demonstrate the formation process of the ellipsoid surface. Through the intuitive demonstration of the teaching aid, it leaves a deep impression on students and improves their learning interest.

[0004] However, deep learning requires a large amount of high-quality labeled data for training and optimization. However, in an educational scenario, it is often difficult to obtain sufficient labeled data. For example, the learning behavior data of students may be incomplete, or the labeling of teaching content lacks consistency. In this case, it will lead to inaccurate prediction of the mastery of knowledge points by certain student groups, thereby affecting the teaching quality and also weakening the professional judgment ability of teachers. Summary of the Invention

[0005] In view of the above problems existing in the existing technical field of data learning and processing, the present invention is proposed.

[0006] Therefore, one of the objectives of the present invention is to provide a course teaching optimization method and system based on deep learning, which can effectively improve the teaching effect and the learning efficiency of students through means such as precise evaluation, personalized teaching, learning behavior optimization, and dynamic adjustment, providing strong support for improving the teaching quality.

[0007] To solve the above technical problems, the present invention provides the following technical solutions:

[0008] On the one hand, the present invention provides a course teaching optimization method based on deep learning, including the following steps:

[0009] Step 1: Obtain the learning data of a preset student, where the learning data includes learning behaviors and learning achievements in a preset subject; and analyze the mastery level of each knowledge point by the preset student based on the learning achievements in the preset subject;

[0010] Step 2: Divide each knowledge point into low-difficulty knowledge points, medium-difficulty knowledge points, and high-difficulty knowledge points according to the mastery situation, and obtain the learning duration of the preset student for the medium-difficulty knowledge points and high-difficulty knowledge points;

[0011] Step 3: During the learning duration, collect the learning characteristics of the preset student for the medium-difficulty knowledge points and high-difficulty knowledge points; the learning characteristics include the knowledge point characteristics of the preset student's learning of the medium-difficulty knowledge points and high-difficulty knowledge points;

[0012] Step 4: Classify the knowledge point characteristics into where n represents the nth knowledge point characteristic; analyze the change in the learning duration of the preset student according to each knowledge point characteristic. When the learning duration of a certain knowledge point characteristic shows a decreasing trend in the future period, it is determined that the preset student has mastered the corresponding knowledge point; otherwise, it is not determined;

[0013] Step 5: During the course teaching process in the future period, generate questions according to the knowledge points mastered by the preset student, obtain the answering time and answering process of the preset student according to the generated questions; obtain the answering ideas of the preset student during the answering process, and evaluate the mastery level of the preset student for the mastered knowledge points according to the answering time.

[0014] As a preferred solution of the present invention, wherein: in the step 1, analyze the mastery level of each knowledge point by the preset student based on the learning achievements in the preset subject, and the analysis method includes analyzing the test paper and / or test content according to the analysis methods of knowledge point scores and high-frequency wrong questions; wherein, the analysis method of the knowledge point scores includes classifying the test paper or test content according to knowledge points and clarifying the examination scope of each knowledge point;

[0015] The analysis method of the high-frequency wrong questions includes finding out the questions with a high error rate in the test paper and analyzing the knowledge points involved in these questions.

[0016] As a preferred solution of the present invention, wherein: in the step 1, the learning behaviors include pre-class learning behaviors, in-class learning behaviors, after-class learning behaviors, and long-term learning behaviors;

[0017] wherein, the pre-class learning behaviors include preview behaviors, knowledge background investigation, and learning plan formulation;

[0018] The classroom learning behaviors include note-taking, classroom interaction, and thinking activity level;

[0019] The after-class learning behaviors include review and consolidation, knowledge expansion, autonomous learning, and learning reflection;

[0020] The long-term learning behaviors include knowledge system construction, learning habit formation, and interdisciplinary learning.

[0021] As a preferred embodiment of the present invention, in step five, according to the answering time, the mastery level of the preset student for the mastered knowledge points is evaluated, including setting a standard answering time, and the setting method includes setting according to the question difficulty and / or the complexity of the knowledge points; it also includes analyzing the difference between the actual answering time of the preset student and the standard answering time; wherein, if the actual answering time of the preset student is lower than the standard answering time and the correct rate is high, it is determined that the preset student has a high mastery level for the corresponding knowledge point;

[0022] If the actual answering time of the preset student is higher than the standard answering time and the correct rate is low, it is determined that the preset student has a low mastery level for the corresponding knowledge point.

[0023] As a preferred embodiment of the present invention, when it is determined that the preset student has a low mastery level for the corresponding knowledge point, the knowledge point is split, and the learning behaviors of the preset student are analyzed according to the split knowledge points, the potential problems of the preset student in a certain knowledge point are predicted, and according to the predicted potential problems, the knowledge points corresponding to the potential problems are emphasized in the future course teaching.

[0024] As a preferred embodiment of the present invention, after analyzing the learning behaviors of the preset student according to the split knowledge points, a course teaching optimization strategy is formulated according to the lack of a certain learning behavior, and the course teaching optimization strategy includes a personalized teaching strategy, a problem situation creation strategy, and an interactive teaching strategy; wherein, the personalized teaching strategy includes adjusting the teaching content and difficulty according to the learning ability and interest of the preset student;

[0025] The problem situation creation strategy includes creating an actual problem situation related to the knowledge point to stimulate the learning interest of students;

[0026] The interactive teaching strategy includes adopting the methods of group discussion and / or interactive Q&A to enhance the classroom participation of students.

[0027] As a preferred embodiment of the present invention, it is provided that: an observation period is preset according to the formulated curriculum teaching optimization strategy, and the observation period is used to observe the mastery level of the preset students on the split knowledge points; within the observation period, the duration of each learning behavior of the preset students is obtained, and the mastery level of the preset students on each split knowledge point within the duration is tested. The testing method includes setting the answering steps of each split knowledge point according to the length of the duration, setting standard answers in each answering step, counting the number of steps correctly answered by the preset students, and calculating the total number of answering steps set for each split knowledge point. If the number of steps correctly answered by the preset students for a certain knowledge point accounts for 80% of the total number, it is determined that the preset students have a high mastery level of the corresponding knowledge point; otherwise, it is not determined.

[0028] As a preferred embodiment of the present invention, it is provided that: when it is determined that the preset students have a low mastery level of the corresponding knowledge point, analyze the differences in the learning behaviors of the preset students on the knowledge points with low mastery level and high mastery level, analyze the associated influence of the differences on the knowledge points with low mastery level of the preset students, and optimize the corresponding learning behaviors according to the associated influence.

[0029] On the other hand, the present invention provides a system applied to a curriculum teaching optimization method based on deep learning, including:

[0030] A data acquisition module, configured to acquire the learning data of preset students, where the learning data includes learning behaviors and learning achievements of a preset subject; and analyze the mastery level of the preset students on each knowledge point based on the learning achievements of the preset subject;

[0031] A data division module, the data division module responds to the data acquisition module, and is configured to divide each knowledge point into low-difficulty knowledge points, medium-difficulty knowledge points, and high-difficulty knowledge points according to the mastery situation, and obtain the learning duration of the preset students on the medium-difficulty knowledge points and high-difficulty knowledge points;

[0032] A data fusion processing unit, the data fusion processing unit responds to the learning duration, and is configured to collect the learning characteristics of the preset students on the medium-difficulty knowledge points and high-difficulty knowledge points during the learning duration; wherein, the learning characteristics include the knowledge point characteristics of the preset students' learning on the medium-difficulty knowledge points and high-difficulty knowledge points; the data fusion processing unit includes a data differentiation module, a determination module, and a data evaluation module;

[0033] The data differentiation module is configured to differentiate the knowledge point characteristics into where n represents the nth knowledge point characteristic; and analyze the change of the learning duration of the preset students according to each knowledge point characteristic;

[0034] When the determination module responds to the change in the learning duration and the learning duration of a certain knowledge point feature shows a decreasing trend in the future period, it is determined that the preset student has mastered the corresponding knowledge point; otherwise, it is not determined.

[0035] The data evaluation module is used to, during the course teaching process in the future period, generate variable questions according to the knowledge points mastered by the preset student, obtain the answering time and answering process of the preset student according to the generated questions; obtain the answering ideas of the preset student during the answering process, and evaluate the mastery level of the preset student on the mastered knowledge points according to the answering time.

[0036] Beneficial effects:

[0037] 1. By analyzing multi-dimensional data such as the learning behavior, learning performance, answering time, and answering process of the preset student, the present invention can accurately evaluate the mastery level of the student on different knowledge points, especially the mastery of medium and high-difficulty knowledge points. This accurate evaluation is beneficial for teachers to more clearly understand the learning status of the preset student.

[0038] 2. By predicting potential problems of the student on certain knowledge points and giving key explanations and interventions in advance, the present invention effectively improves the pertinence and efficiency of teaching and realizes individualized teaching.

[0039] 3. The present invention not only pays attention to the mastery level of knowledge points, but also formulates personalized learning behavior optimization strategies for the preset student by analyzing the learning behavior (such as learning duration, learning characteristics, etc.) of the preset student; for example, by creating problem situations, interactive teaching, etc., to stimulate the learning interest and participation of the preset student and help the preset student develop good learning habits.

[0040] 4. By setting an observation period, the present invention monitors the learning behavior and knowledge point mastery of the preset student after the implementation of the optimization strategy; if it is found that the mastery level of the preset student on certain knowledge points is still low, the system will further analyze the reasons and dynamically adjust the learning behavior optimization strategy to ensure the continuous improvement of teaching effects. Brief description of the drawings

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them:

[0042] Figure 1 It is a modular structure schematic diagram of the course teaching optimization system based on deep learning according to the embodiment of the present invention;

[0043] Figure 2 It is a schematic flowchart of the method according to an embodiment of the present invention;

[0044] Reference numerals in the figure: 110 - data acquisition module; 120 - data division module; 130 - data fusion processing unit; 1301 - data differentiation module; 1302 - determination module; 1303 - data evaluation module. Specific implementation manner

[0045] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of the present invention fall within the scope of protection of the present invention.

[0046] Since deep learning requires a large amount of high-quality labeled data for training and optimization, but in an educational scenario, it is often difficult to obtain sufficient labeled data. For example, the learning behavior data of students may be incomplete, or the labeling of teaching content lacks consistency. In this case, it will lead to inaccurate prediction of the mastery of knowledge points by certain student groups, thereby affecting the teaching quality and also weakening the professional judgment ability of teachers.

[0047] Based on this, the present invention proposes a curriculum teaching optimization method and system based on deep learning, which can effectively improve the teaching effect and the learning efficiency of students through means such as accurate evaluation, personalized teaching, learning behavior optimization, and dynamic adjustment, providing strong support for improving the teaching quality.

[0048] The following further specifically describes this solution through embodiments and in conjunction with the accompanying drawings.

[0049] Refer to Figures 1 to 2 , which is an embodiment of the present invention. This embodiment provides a curriculum teaching optimization method based on deep learning, including the following steps:

[0050] Step 1: Obtain the learning data of a preset student, where the learning data includes learning behavior and the learning achievements of a preset subject; and analyze the mastery degree of each knowledge point of the preset student based on the learning achievements of the preset subject;

[0051] It should be noted in this embodiment that the analysis method includes analyzing the test paper and / or test content according to the analysis methods of knowledge point scores and high-frequency wrong questions; among them, the analysis method of knowledge point scores includes classifying the test paper or test content according to knowledge points to clarify the examination scope of each knowledge point;

[0052] The analysis methods for high-frequency wrong questions include finding out the questions with high error rates in the test papers and analyzing the knowledge points involved in these questions;

[0053] It should also be specifically stated in this embodiment that learning behaviors include pre-class learning behaviors, in-class learning behaviors, after-class learning behaviors, and long-term learning behaviors;

[0054] Among them, pre-class learning behaviors include preview behaviors, knowledge background investigation, and learning plan formulation;

[0055] In-class learning behaviors include note-taking, in-class interaction, and thinking activity;

[0056] After-class learning behaviors include review and consolidation, knowledge expansion, autonomous learning, and learning reflection;

[0057] Long-term learning behaviors include knowledge system construction, learning habit formation, and interdisciplinary learning;

[0058] Step 2: According to the mastery situation, divide each knowledge point into low-difficulty knowledge points, medium-difficulty knowledge points, and high-difficulty knowledge points, and obtain the learning duration of the preset students for medium-difficulty knowledge points and high-difficulty knowledge points;

[0059] Step 3: In the learning duration, collect the learning characteristics of the preset students for medium-difficulty knowledge points and high-difficulty knowledge points; The learning characteristics include the knowledge point characteristics of the preset students' learning of medium-difficulty knowledge points and high-difficulty knowledge points;

[0060] Step 4: Divide the knowledge point characteristics into Among them, n represents the nth knowledge point characteristic; Analyze the change of the learning duration of the preset students according to each knowledge point characteristic. When the learning duration of a certain knowledge point characteristic shows a decreasing trend in the future period, it is determined that the preset students have mastered the corresponding knowledge point; Otherwise, it is not determined;

[0061] Step 5: In the course teaching process in the future period, formulate questions according to the knowledge points mastered by the preset students, obtain the answering time and answering process of the preset students according to the formulated questions; Obtain the answering ideas of the preset students during the answering process, and evaluate the mastery degree of the preset students for the mastered knowledge points according to the answering time;

[0062] It should be noted in this embodiment that the question types for formulating questions include calculation questions, discussion questions, and proof questions;

[0063] In this embodiment, it should be noted that according to the answering time, the mastery level of the preset students' mastered knowledge points is evaluated, including setting the standard answering time, and the setting method includes setting according to the question difficulty and / or the complexity of the knowledge points; it also includes analyzing the difference between the actual answering time of the preset students and the standard answering time; among them, if the actual answering time of the preset students is lower than the standard answering time and the correct rate is high, it is determined that the preset students have a high mastery level of the corresponding knowledge points;

[0064] If the actual answering time of the preset students is higher than the standard answering time and the correct rate is low, it is determined that the preset students have a low mastery level of the corresponding knowledge points;

[0065] On this basis, when it is determined that the preset students have a low mastery level of the corresponding knowledge points, the knowledge points are split, and the learning behaviors of the preset students are analyzed according to the split knowledge points, the potential problems of the preset students in a certain knowledge point are predicted, and according to the predicted potential problems, the knowledge points corresponding to the potential problems are focused on in the future course teaching;

[0066] It should be noted that in this embodiment, the key teaching methods include step-by-step teaching and gradual practice, repeated practice and intensive training, and summary and knowledge framework;

[0067] Among them, the method of step-by-step teaching and gradual practice includes decomposing complex knowledge points into multiple small steps, teaching step by step and arranging practice, thereby reducing the learning difficulty and helping students gradually master the knowledge;

[0068] Repeated practice and intensive training can help students consolidate weak knowledge points and enhance their memory and understanding ability;

[0069] Summary and knowledge framework can help students summarize knowledge points, build a knowledge framework, enhance the systematicness of knowledge, thereby helping students understand knowledge points as a whole and improve learning efficiency;

[0070] In this embodiment, further, after analyzing the learning behaviors of the preset students according to the split knowledge points, a course teaching optimization strategy is formulated according to the lack of a certain learning behavior. The course teaching optimization strategy includes personalized teaching strategy, creating problem situation strategy and interactive teaching strategy; among them, the personalized teaching strategy includes adjusting the teaching content and difficulty according to the learning ability and interest of the preset students;

[0071] For example, for students with relatively weak learning ability, more basic exercises and tutoring can be provided;

[0072] The creating problem situation strategy includes creating an actual problem situation related to the knowledge points to stimulate students' learning interest;

[0073] For example, when explaining "chemical reaction rate", actual industrial production problems can be incorporated;

[0074] Interactive teaching strategies include using group discussions and / or interactive Q&A methods to enhance students' sense of classroom participation;

[0075] For example, when explaining "mathematical functions", students can be divided into groups to discuss the properties of function graphs;

[0076] Specifically in this embodiment, an observation period is preset according to the formulated curriculum teaching optimization strategy. The observation period is used to observe the mastery level of the preset students for the split knowledge points. During the observation period, the duration of each learning behavior of the preset students is obtained, and the mastery level of the preset students for each split knowledge point within the duration is tested. The testing method includes setting the answering steps for each split knowledge point according to the length of the duration, setting standard answers in each answering step, counting the number of steps correctly answered by the preset students, and calculating the total number of answering steps set for each split knowledge point. If the number of steps correctly answered by the preset students for a certain knowledge point accounts for 80% of the total number, it is determined that the preset students have a high mastery level for the corresponding knowledge point; otherwise, it is not determined.

[0077] It should be noted that in this embodiment, the observation period includes 7 to 15 natural days;

[0078] Further in this embodiment, when it is determined that the preset students have a low mastery level for the corresponding knowledge point, analyze the differences in the learning behaviors of the preset students for the knowledge points with low mastery level and the knowledge points with high mastery level, analyze the associated influence of the differences on the knowledge points with low mastery level of the preset students, and optimize the corresponding learning behaviors according to the associated influence.

[0079] It should be noted that optimizing the corresponding learning behaviors according to the associated influence includes adjusting the personalized learning path for the preset students and customizing a personalized learning path for them. For example, for knowledge points with low mastery level, more basic exercises and / or video explanations can be recommended to the preset students.

[0080] Based on the above, by analyzing multi-dimensional data such as the learning behaviors, learning achievements, answering time, and answering process of the preset students, this application can accurately evaluate the mastery level of the preset students for different knowledge points, especially for medium and high-difficulty knowledge points. This accurate evaluation is conducive to teachers understanding the learning status of the preset students more clearly, thereby ensuring the improvement of teaching effects.

[0081] Combined with the above curriculum teaching optimization method based on deep learning, this embodiment also proposes the working system of this method as follows:

[0082] A data acquisition module 110, which is used to acquire the learning data of a preset student. The learning data includes learning behaviors and learning achievements in a preset subject; and based on the learning achievements in the preset subject, analyze the mastery degree of each knowledge point by the preset student;

[0083] A data division module 120. The data division module responds to the data acquisition module and is used to divide each knowledge point into low-difficulty knowledge points, medium-difficulty knowledge points, and high-difficulty knowledge points according to the mastery situation, and obtain the learning duration of the preset student for medium-difficulty knowledge points and high-difficulty knowledge points;

[0084] A data fusion processing unit 130. The data fusion processing unit responds to the learning duration and is used to collect the learning characteristics of the preset student for medium-difficulty knowledge points and high-difficulty knowledge points in the learning duration; wherein, the learning characteristics include the knowledge point characteristics of the preset student's learning of medium-difficulty knowledge points and high-difficulty knowledge points. The data fusion processing unit 130 includes a data differentiation module 1301, a determination module 1302, and a data evaluation module 1303;

[0085] The data differentiation module 1301 is used to distinguish the knowledge point characteristics into wherein, n represents the nth knowledge point characteristic; and analyze the change of the learning duration of the preset student according to each knowledge point characteristic;

[0086] The determination module 1302 responds to the change of the learning duration. When the learning duration of a certain knowledge point characteristic shows a decreasing trend in the future period, it is determined that the preset student has mastered the corresponding knowledge point; otherwise, it is not determined;

[0087] The data evaluation module 1303 is used to, during the course teaching process in the future period, generate variable questions according to the knowledge points mastered by the preset student, obtain the answering time and answering process of the preset student according to the generated questions; obtain the answering ideas of the preset student during the answering process, and evaluate the mastery degree of the preset student for the mastered knowledge points according to the answering time.

[0088] In summary, through means such as precise evaluation, personalized teaching, learning behavior optimization, and dynamic adjustment, the present invention can effectively improve the teaching effect and the learning efficiency of students, providing strong support for improving the teaching quality.

[0089] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A course teaching optimization method based on deep learning, characterized in that: The following steps are involved: Step 1: Obtaining learning data of a preset student, wherein the learning data includes learning behavior and learning performance of a preset subject; and analyzing the mastery of each knowledge point by the preset students based on the learning performance of the preset subjects; Step 2: Divide each knowledge point into low-difficulty knowledge points, medium-difficulty knowledge points and high-difficulty knowledge points according to the mastery status, and obtain the learning time of the preset students for the medium-difficulty knowledge points and the high-difficulty knowledge points; Step 3: During the learning time, the learning characteristics of the preset students on the medium-difficulty knowledge points and the high-difficulty knowledge points are collected; the learning characteristics include the knowledge point characteristics of the preset students on the medium-difficulty knowledge points and the high-difficulty knowledge points; Step 4: Distinguish the knowledge point features into Among them, n represents the feature of the nth knowledge point; Analyze the changes in the learning time of the preset students according to the characteristics of each knowledge point. When the learning time of a certain knowledge point characteristic in the future period shows a decreasing trend, it is determined that the preset students have mastered the corresponding knowledge point; otherwise, it is not determined; Step 5: In the course teaching process in the future period, change the proposed questions according to the knowledge points that the preset students have mastered, and obtain the answering time and answering process of the preset students according to the proposed questions; obtain the answering ideas of the preset students in the answering process, and evaluate the mastery of the preset students on the mastered knowledge points according to the answering time.

2. A course teaching optimization method based on deep learning as claimed in claim 1, characterized in that: In the step 1, the mastery of each knowledge point of the preset student is analyzed based on the learning performance of the preset subject, and the analysis method includes analyzing the test paper and / or test content according to the knowledge point score and the analysis method of high-frequency wrong questions; wherein the analysis method of the knowledge point score includes classifying the test paper or test content according to the knowledge point and clarifying the examination scope of each knowledge point; The method for analyzing the high-frequency wrong questions includes finding out the questions with high error rates in the test paper and analyzing the knowledge points involved in these questions.

3. A course teaching optimization method based on deep learning as claimed in claim 1, characterized in that: In the step 1, the learning behavior includes pre-class learning behavior, class learning behavior, after-class learning behavior and long-term learning behavior; The pre-class learning behaviors include previewing, background knowledge investigation, and study plan formulation; The classroom learning behaviors include note-taking, classroom interaction, and active thinking; The after-class learning behaviors include review and consolidation, knowledge expansion, independent learning and learning reflection; The long-term learning behavior includes knowledge system construction, learning habit formation and interdisciplinary learning.

4. A course teaching optimization method based on deep learning as claimed in claim 1, characterized in that: In the step 5, based on the answering time, the mastery of the knowledge points mastered by the preset students is evaluated, including setting a standard answering time, wherein the setting method includes setting it according to the difficulty of the question and / or the complexity of the knowledge point; and also includes analyzing the difference between the actual answering time of the preset students and the standard answering time; wherein, if the actual answering time of the preset students is lower than the standard answering time and the accuracy rate is high, it is determined that the preset students have a high mastery of the corresponding knowledge points; If the actual answering time of the preset student is higher than the standard answering time and the accuracy rate is low, it is judged that the preset student has a low level of mastery of the corresponding knowledge points.

5. A course teaching optimization method based on deep learning as claimed in claim 4, characterized in that: When it is determined that the preset student has a low level of mastery of the corresponding knowledge points, the knowledge points are split, and the learning behavior of the preset student is analyzed based on the split knowledge points, and potential problems of the preset student in one of the knowledge points are predicted. Based on the predicted potential problems, the knowledge points corresponding to the potential problems are focused on in the course teaching in the future period.

6. A course teaching optimization method based on deep learning as claimed in claim 5, characterized in that: After analyzing the learning behavior of the preset students according to the split knowledge points, a course teaching optimization strategy is formulated according to the lack of a certain learning behavior, and the course teaching optimization strategy includes a personalized teaching strategy, a problem situation creation strategy and an interactive teaching strategy; wherein the personalized teaching strategy includes adjusting the teaching content and difficulty according to the learning ability and interest of the preset students; The strategy of creating problem situations includes creating actual problem situations related to knowledge points to stimulate students' learning interest; The interactive teaching strategy includes adopting group discussion and / or interactive question-and-answer methods to enhance students' sense of classroom participation.

7. A course teaching optimization method based on deep learning as claimed in claim 6, characterized in that: An observation period is preset according to the formulated course teaching optimization strategy, and the observation period is used to observe the mastery of the split knowledge points by the preset students; during the observation period, the duration of each learning behavior of the preset students is obtained, and the mastery of the split knowledge points by the preset students within the duration is tested, and the testing method includes setting the solution steps of each split knowledge point according to the length of the duration, and setting a standard answer in each solution step, counting the number of steps correctly answered by the preset students, and calculating the total number of solution steps set for each split knowledge point. If the number of steps correctly answered by the preset students for a certain knowledge point accounts for 80% of the total number, then it is determined that the preset students have a high degree of mastery of the corresponding knowledge point; otherwise, no determination is made.

8. A course teaching optimization method based on deep learning as claimed in claim 7, characterized in that: When it is determined that the preset student has a low level of mastery of the corresponding knowledge points, the difference in learning behaviors of the preset students for the knowledge points with a low level of mastery and the knowledge points with a high level of mastery is analyzed, the associated influence of the difference and the knowledge points with a low level of mastery by the preset students is analyzed, and the corresponding learning behavior is optimized according to the associated influence.

9. A system applied to a course teaching optimization method based on deep learning as described in claims 1 to 8, characterized in that: include: A data acquisition module, used to acquire learning data of a preset student, wherein the learning data includes learning behavior and learning performance of a preset subject; and analyzing the mastery of each knowledge point by the preset students based on the learning performance of the preset subjects; A data division module, the data division module responds to the data acquisition module, and is used to divide each knowledge point into low-difficulty knowledge points, medium-difficulty knowledge points and high-difficulty knowledge points according to the mastery status, and obtain the learning time of the preset students for the medium-difficulty knowledge points and the high-difficulty knowledge points; A data fusion processing unit, the data fusion processing unit responds to the learning time, and is used to collect the learning characteristics of the preset students on the medium-difficulty knowledge points and the high-difficulty knowledge points during the learning time; wherein the learning characteristics include the knowledge point characteristics of the preset students learning the medium-difficulty knowledge points and the high-difficulty knowledge points; the data fusion processing unit includes a data differentiation module, a determination module and a data evaluation module; The data distinguishing module is used to distinguish the knowledge point features into Wherein, n represents the feature of the nth knowledge point; and the change of the preset student learning time is analyzed according to the features of each knowledge point; The determination module responds to the change of the learning time, and when the learning time of a certain knowledge point feature in the future period shows a decreasing trend, it is determined that the preset student has mastered the corresponding knowledge point; otherwise, no determination is made; The data evaluation module is used to change the proposed questions according to the knowledge points that the preset students have mastered during the course teaching process in the future period, and obtain the answering time and answering process of the preset students according to the proposed questions; obtain the answering ideas of the preset students in the answering process, and evaluate the mastery of the preset students on the mastered knowledge points based on the answering time.

Images