Intelligent education system based on multi-algorithm fusion

Abstract

The invention relates to the technical field of intelligent education, in particular to an intelligent education system based on multi-algorithm fusion. The system comprises a processing analysis unit, an extraction construction unit and a personalized planning unit. The thought adaptation module performs knowledge reasoning of various problem solving thoughts according to the constructed thought model data, and then performs matching according to the reasoned problem solving thought data and problem solving behaviors in the behavior model data to obtain a problem solving thought with the highest probability of the problem solving behaviors of the students; different students have different learning styles and ability levels, so that the learning effect and the learning motivation can be improved by recommending the most suitable problem solving thought according to the problem solving behaviors of the students, meanwhile, an easily understood problem solving thought is recommended for the students to select, a simple and easily understood problem solving thought is recommended, the students are helped to understand and think problems, and the learning effect and the learning motivation are improved. The simple problem solving thought can tease the main logic of the problem, and helps students concentrate attention and grasp the problem solving thought better.

Description

Technical Field

[0001] The present invention relates to the field of intelligent education technology, and in particular to an intelligent education system based on multi-algorithm fusion. Background Art

[0002] In real scenarios, intelligent education systems play an important role in our lives. Intelligent education systems can be widely used in personalized learning support, teaching resource recommendation, learning progress monitoring, intelligent evaluation and feedback, teaching decision support and intelligent teaching assistance. They can comprehensively improve the level of education and teaching, and promote students' personalized learning and all-round development. When students raise questions that they do not understand, the general intelligent education system only explains the questions and difficulties, and provides a difficult-to-understand solution. If the student has a solid knowledge base, he can understand the solution provided. If the student has a weak knowledge base or is not used to this solution, the student may not understand all of it. Although he can understand the solution in time, the student is not good at it. Therefore, the error rate of using the provided solution is high. Therefore, in order to avoid the general intelligent education system providing a difficult-to-understand solution and a problem-solving behavior habit that is not suitable for the student, we provide an intelligent education system based on multi-algorithm fusion. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent education system based on multi-algorithm fusion to solve the problems raised in the above background technology.

[0004] To achieve the above-mentioned purpose, the present invention provides an intelligent education system based on multi-algorithm fusion, including a processing and analysis unit, an extraction and construction unit, and a personalized planning unit;

[0005] The processing and analysis unit is used to obtain and process the student's relevant learning data and other historical data, and then analyze the processed data to analyze the student's learning behavior patterns, difficult problem-solving ideas and psychological emotions;

[0006] The extraction and construction unit is used to receive the data in the processing and analysis unit and extract relevant data and construct a model;

[0007] The personalized planning unit is used to receive the data in the processing and analysis unit and extract the model data constructed in the construction unit to perform personalized recommendation and learning path planning;

[0008] The extraction and construction unit receives the data in the processing and analysis unit, and extracts relevant data based on the learning difficulty problem data and the student learning rules analyzed in the processing and analysis unit, and then constructs a thinking model and a behavior model based on the extracted relevant data, and transmits the constructed thinking model and behavior model data to the personality planning unit. The personality planning unit makes personalized thinking recommendations and personalized learning recommendations based on the thinking model and behavior model data constructed in the extraction and construction unit and the data processed in the processing and analysis unit, and then plans the learning path based on the personalized learning recommendation data and the psychological data analyzed in the processing and analysis unit.

[0009] As a further improvement of the technical solution, the processing and analysis unit includes an acquisition processing module, a data analysis module, and a data evaluation module;

[0010] The acquisition and processing module is used to obtain student-related learning data and other historical data, the acquired student-related learning data includes learning conditions, difficult questions raised by students and psychological emotional data, and then process the acquired student-related learning data;

[0011] Learning data include students’ academic performance, learning behaviors, learning styles, and preferences;

[0012] The data analysis module is used to receive the data processed in the acquisition processing module, and use the clustering analysis algorithm to analyze the students' learning rules based on the processed learning situation data, and then use the knowledge solution algorithm to analyze the difficult problems based on the difficult problems raised by the processed students and the relevant knowledge points of the difficult problems in the database, use the hierarchical clustering algorithm to perform psychological analysis based on the processed psychological and emotional data, and then use the multi-algorithm fusion technology to fuse the data analyzed by the clustering analysis algorithm, the data analyzed by the knowledge solution algorithm and the data analyzed by the hierarchical clustering algorithm into a data set.

[0013] As a further improvement of the present technical solution, the data evaluation module is used to receive the student-related learning data processed in the acquisition processing module and the student learning regularity data of the data set in the data analysis module, and use the shortcoming evaluation technology to evaluate the student's knowledge shortcomings based on the processed student-related learning data and the student learning regularity data of the data set, and evaluate the deficiencies in the student's understanding of a certain subject knowledge during the learning process.

[0014] As a further improvement of the technical solution, the extraction construction unit includes a model construction module and a thought adaptation module;

[0015] The model construction module is used to receive the student learning rule data and the analyzed difficult problem-solving ideas data in the data set in the data analysis module and extract the students' relevant learning rule data and difficult problem-solving ideas data, and use the training model technology to construct a behavior model based on the extracted student learning rule data, and then use the training model technology to construct an idea model based on the extracted difficult problem-solving ideas data.

[0016] As a further improvement of the present technical solution, the idea adaptation module is used to receive the idea model data and the behavior model data constructed in the model construction module, and use the knowledge graph technology to perform knowledge reasoning of multiple problem-solving ideas based on the constructed idea model data, and then match the problem-solving behaviors in the inferred multiple problem-solving idea data and behavior model data to match a problem-solving idea with the highest probability of the student's problem-solving behavior, and at the same time recommend an easy-to-understand problem-solving idea for students to choose.

[0017] As a further improvement of the present technical solution, the personalized planning unit includes a personalized recommendation module and a learning path planning module;

[0018] The personalized recommendation module is used to receive the idea model data in the model construction module, the matching problem-solving idea data in the idea adaptation module and the data processed in the acquisition processing module, and use the learning recommendation technology to make personalized learning recommendations based on the idea model data, the processed data and the matching problem-solving idea data. The personalized recommendations include suitable learning resources, course content and exercise questions.

[0019] As a further improvement of the present technical solution, the learning path planning module is used to receive the personalized learning data recommended in the personalized recommendation module, obtain the data processed in the processing module and the sentiment data of the data set in the data analysis module, and use planning decision-making technology to plan the learning path based on the recommended personalized learning data, the processed data and the sentiment data of the data set.

[0020] As a further improvement of this technical solution, the planning decision technology implementation principle in the learning path planning module is:

[0021] Principle ①: First, we need to clarify the learner's planning goals and needs, including the learning topic, difficulty, time and other requirements, and collect the recommended personalized learning data, processed data and emotional data of the data set;

[0022] Principle ②: Use decision trees to build a planning model based on the recommended personalized learning data, processed data, and sentiment data of the data set;

[0023] Principle ③: Establish a learning knowledge graph based on the planning model, and then plan a suitable learning path based on the established learning knowledge graph data.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] In this multi-algorithm integrated intelligent education system, the idea adaptation module performs knowledge reasoning of multiple problem-solving ideas based on the constructed idea model data, and then matches the problem-solving behaviors in the inferred multiple problem-solving idea data and the behavior model data to match a problem-solving idea with the highest probability of the student's problem-solving behavior. Different students have different learning styles and ability levels. Therefore, recommending the most appropriate problem-solving idea based on the student's problem-solving behavior can improve learning effects and learning motivation. At the same time, recommending an easy-to-understand problem-solving idea for students to choose. Recommending a concise and easy-to-understand problem-solving idea helps students understand and think about the problem. Concise problem-solving ideas can often sort out the main logic and key points of the problem, help students concentrate and better master problem-solving ideas. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is an overall block diagram of the present invention;

[0027] Figure 2 It is a block diagram of the processing and analysis unit of the present invention;

[0028] Figure 3 A block diagram of the extraction construction unit of the present invention;

[0029] Figure 4 It is a block diagram of the personalized planning unit of the present invention.

[0030] The meaning of each number in the figure is:

[0031] 1. Processing and analysis unit; 11. Acquisition and processing module; 12. Data analysis module; 13. Data evaluation module;

[0032] 2. Extraction construction unit; 21. Model construction module; 22. Idea adaptation module;

[0033] 3. Personalized planning unit; 31. Personalized recommendation module; 32. Learning path planning module. DETAILED DESCRIPTION

[0034] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Example 1

[0036] Considering that when students raise questions that contain incomprehensible doubts and difficulties, since general intelligent education systems only provide corresponding explanations for the doubts and difficulties and provide a difficult-to-understand solution, if the student has a solid knowledge base, he or she can understand the provided solution. If the student has a weak knowledge base or is not used to this solution, the student may not understand all of it. Even if the student understands the solution, the error rate when using the provided solution is high because the student is not good at it. Therefore, in order to avoid the general intelligent education system providing a difficult-to-understand solution and a problem-solving behavior habit that is not suitable for the student, the present invention provides an intelligent education system based on multi-algorithm fusion, please refer to Figure 1-Figure 4 , including a processing and analysis unit 1, an extraction and construction unit 2, and a personality planning unit 3;

[0037] The processing and analysis unit 1 is used to obtain and process the student's relevant learning data and other historical data, and then analyze the processed data for the student's learning behavior patterns, difficult problem-solving ideas and psychological emotions. The extraction and construction unit 2 is used to receive the data in the processing and analysis unit 1 and extract relevant data and construct a model. The personalized planning unit 3 is used to receive the data in the processing and analysis unit 1 and the model data constructed in the extraction and construction unit 2 for personalized recommendation and learning path planning;

[0038] The extraction and construction unit 2 receives the data in the processing and analysis unit 1, and extracts relevant data based on the learning difficulty problem data and the student learning rules analyzed in the processing and analysis unit 1, and then constructs a thought model and a behavior model based on the extracted relevant data, and transmits the constructed thought model and behavior model data to the personality planning unit 3. The personality planning unit 3 makes personalized thought recommendations and personalized learning recommendations based on the thought model and behavior model data constructed in the extraction and construction unit 2 and the data processed in the processing and analysis unit 1, and then plans the learning path based on the personalized learning recommendation data and the psychological data analyzed in the processing and analysis unit 1. This allows students to learn more specifically and avoid anxiety caused by learning content that is too fragmented or difficult to understand. Anxiety during the learning process may affect students' learning effects, and personalized learning paths can help alleviate this anxiety.

[0039] The following is a refinement of the above units, see Figure 2-Figure 4 ;

[0040] The processing and analysis unit 1 includes an acquisition processing module 11, a data analysis module 12, and a data evaluation module 13;

[0041] The acquisition and processing module 11 is used to obtain student-related learning data and other historical data. The acquired student-related learning data includes learning status, difficult questions raised by students and psychological emotional data. The learning status data includes students' academic performance, learning behavior, learning style and preferences. The acquired student-related learning data is then processed to reduce errors in the data by processing the acquired data.

[0042] The data analysis module 12 is used to receive the data processed in the acquisition processing module 11, and use the cluster analysis algorithm to analyze the learning rules of students based on the processed learning situation data, and then use the knowledge solution algorithm to analyze the difficult problems raised by the processed students and the relevant knowledge points of the difficult problems in the database, and use the hierarchical clustering algorithm to perform psychological analysis based on the processed psychological and emotional data, and then use the multi-algorithm fusion technology to fuse the data analyzed by the cluster analysis algorithm, the data analyzed by the knowledge solution algorithm and the data analyzed by the hierarchical clustering algorithm into a data set;

[0043] Cluster analysis algorithm implementation principle:

[0044] Principle 1: According to the characteristics and data scale of the learning data being processed, the selection of clustering algorithm is determined according to the data characteristics and application requirements. Different algorithms have different clustering principles and applicable scenarios;

[0045] Principle 2: Before applying the clustering algorithm, it is necessary to determine the number of clusters, that is, how many categories the students are divided into. In the student learning data, the number of student groups needs to be determined based on the actual situation and analysis purpose;

[0046] Principle 3: Group students based on similarity measures between data. By calculating the similarity measures between students’ data, similar students can be grouped into the same category.

[0047] Principle 4: Students are gradually grouped according to similarity measures until the set number of clusters and clustering criteria are met. Each cluster represents a group of students with similar learning characteristics and rules. The student learning rules are analyzed based on the learning characteristics and rules.

[0048] Knowledge solution algorithm implementation steps:

[0049] Step ①: First, we need to deeply understand the difficult questions raised by students and the difficult questions in the database, and ensure that we have a clear understanding of the background and key points of the questions. Determining the difficulty is the first step in understanding the problem and solving it. On this basis, we can establish ideas for solving the problem;

[0050] Step ②: For difficult questions, you need to determine the relevant knowledge points and concepts involved. For questions raised by students, you can find relevant knowledge points through textbooks, materials and other resources; for database problems, you can consult database theory and related technical literature. Finding relevant knowledge points is the key to solving problems and can help establish the foundation and methodology for solving problems;

[0051] Step 3: According to the difficult questions and related knowledge points, sort out the problem-solving ideas and determine the logical order and steps of the problem-solving. The problem-solving ideas can be broken down into several key steps to ensure that the problem is solved step by step;

[0052] Step ④: Based on the knowledge points learned and the problem-solving ideas sorted out, gradually solve difficult problems. In this process, you can combine theoretical knowledge and practical operations to find appropriate problem-solving methods.

[0053] Implementation principle of hierarchical clustering algorithm:

[0054] Principle ①, first of all, it is necessary to collect and process psychological and emotional data to ensure that the data meets the input requirements of the hierarchical clustering algorithm;

[0055] Principle ②: Before performing hierarchical clustering, it is necessary to define a suitable similarity metric to measure the similarity between different sentiment data samples;

[0056] Principle ③: Based on the defined similarity measure, similar sentiment samples are gradually merged in the hierarchical clustering algorithm to form a cluster hierarchy, which means that each sample is initially regarded as a separate cluster, and then gradually merged into larger clusters based on similarity until the entire data set is merged into a large cluster;

[0057] Principle ④: By analyzing the clustering results, the relationships and patterns between different emotions can be revealed. The clustering results can be represented as a dendrogram (clustering tree). The similarities and differences between samples can be understood based on the structure of the dendrogram and the distance between nodes.

[0058] Principle ⑤: Based on the clustering results, psychological analysis can be performed to discover the associations between different emotion categories, the changing trends of emotional states, the characteristics of emotional patterns, and to understand the psychological and emotional states and changes of individuals or groups, and to obtain the final psychological and emotional state data.

[0059] Multi-algorithm fusion technology solid steps:

[0060] Step 1: First, ensure that each algorithm can provide the required analysis results. Each algorithm should provide its corresponding analysis results, including the student learning regularity data analyzed by the clustering analysis algorithm, the problem-solving ideas data of the difficult problems analyzed by the knowledge solution algorithm, and the psychological emotion data analyzed by the hierarchical clustering algorithm;

[0061] Step 2: Add corresponding labels to each analyzed data sample. The labels can be added based on the classification of regular clustering clusters, the topic category of knowledge answers, and the sentiment category of sentiment analysis;

[0062] Step 3: Fusion the data results from different algorithms according to the corresponding features and labels, and build a unified data set.

[0063] The data evaluation module 13 is used to receive the student-related learning data processed in the acquisition processing module 11 and the student learning regularity data of the data set in the data analysis module 12, and use the short board evaluation technology to evaluate the student's knowledge shortcomings according to the processed student-related learning data and the student learning regularity data of the data set, and evaluate the deficiencies of the student's understanding of a certain subject knowledge during the learning process. By evaluating the deficiencies of the student's understanding of a certain subject knowledge, the educator can provide personalized learning guidance for the student, strengthen the learning of relevant knowledge points in a targeted manner, and help the student fill the knowledge gap;

[0064] Steps to implement the short board assessment technology:

[0065] Step 1: First, calculate the student's score on each knowledge point based on the processed student-related learning data and the student learning regularity data in the data set;

[0066] Step 2: Record the output results to get the students' scores on each knowledge point. These scores can be used to preliminarily judge the students' mastery of each knowledge point.

[0067] Step 3. By analyzing students' scores on various knowledge points, we can determine which knowledge points have shortcomings or learning difficulties, and based on the set evaluation criteria, determine which knowledge points students' scores are lower than expected, thereby determining which knowledge points are students' shortcomings.

[0068] The extraction construction unit 2 includes a model construction module 21 and a thought adaptation module 22;

[0069] The model building module 21 is used to receive the student learning regularity data and the analyzed difficult problem-solving ideas data of the data set in the data analysis module 12, and extract the relevant learning regularity data and difficult problem-solving ideas data of the students, and use the training model technology to build a behavior model based on the extracted student learning regularity data, and then use the training model technology to build a thinking model based on the extracted difficult problem-solving ideas data (the implementation steps of the constructed thinking model and the constructed behavior model are the same, and the extracted student learning regularity data are replaced by the extracted difficult problem-solving ideas data);

[0070] Steps to build a behavioral model:

[0071] Step ①, obtain the extracted student learning regularity data, the parameters of the selected model can be started by random initialization, and then the optimizer is used to update the parameters so that the loss function gradually decreases;

[0072] Step ②: Divide the extracted student learning regularity dataset into a training set and a test set. The training set data is used to train the model, and the validation set data is used to evaluate the performance of the model. When the training results of the model overfit the training data, the performance on the validation set will be poor. Therefore, before training the model, the data needs to be randomized and divided.

[0073] Step 3: According to the needs of the problem and the characteristics of the data, select a suitable loss function and conduct training, perform continuous iterative training, and stop training when the training data reaches the set target data, and finally build a behavioral model.

[0074] The idea adaptation module 22 is used to receive the idea model data and the behavior model data constructed in the model construction module 21, and use the knowledge graph technology to perform knowledge reasoning of multiple problem-solving ideas based on the constructed idea model data, and then match the problem-solving behaviors in the reasoned multiple problem-solving idea data and the behavior model data to match a problem-solving idea with the highest probability of the student's problem-solving behavior, which refers to the problem-solving idea that the student uses most often, and at the same time recommend an easy-to-understand problem-solving idea for the student to choose;

[0075] Implementation principle of knowledge graph technology:

[0076] Principle 1: First, build a knowledge graph that contains the relationship between various questions and solution ideas. The knowledge graph simulates the relationship between entities and concepts in the real world and represents knowledge in the form of triples.

[0077] Principle 2: Convert problem-solving ideas and knowledge points from various sources into structured triples, and continuously expand and improve the knowledge graph to include more problem-solving ideas and related knowledge points;

[0078] Principle 3: Use the association relationships and semantic information in the graph to perform reasoning, integrate information from different knowledge graph nodes and relationships, and use multi-source information for comprehensive reasoning. For a specific question, combine relevant knowledge points, problem-solving ideas and question requirements to find relevant problem-solving ideas from the knowledge graph and perform knowledge reasoning on multiple problem-solving ideas.

[0079] The personalized planning unit 3 includes a personalized recommendation module 31 and a learning path planning module 32;

[0080] The personalized recommendation module 31 is used to receive the idea model data in the model building module 21, the problem-solving idea data matched in the idea adaptation module 22, and the data processed in the acquisition processing module 11, and use the learning recommendation technology to make personalized learning recommendations based on the idea model data, the processed data, and the matched problem-solving idea data. The recommended personalization includes suitable learning resources, course content, and exercise questions;

[0081] At the same time, the personalized recommendation module 31 also receives the student knowledge short board data evaluated in the data evaluation module 13 and the behavior model data in the model construction module 21, and uses the learning recommendation technology to make personalized learning recommendations based on the behavior model data and the processed data and the evaluated student knowledge short board data;

[0082] Learning recommendation technology implementation principle:

[0083] Principle ①: Collect and process student data, problem-solving ideas data, and learning behavior model data, which may include students' learning resource browsing records, problem-solving records, study notes, and test scores;

[0084] Principle ②: Convert the problem-solving ideas into numerical feature representations, and process the students’ processing data and behavior model data, which may include vectorizing the processing data and behavior model data;

[0085] Principle ③: Measure the similarity relationship between the behavior model data and the problem-solving idea data based on similarity. Based on the similarity results and combined with the students' personal learning characteristics and preferences, determine the most suitable personalized learning recommendations for the students.

[0086] The learning path planning module 32 is used to receive the personalized learning data recommended by the personalized recommendation module 31, obtain the data processed by the processing module 11 and the sentiment data of the data set in the data analysis module 12, and use the planning decision technology to plan the learning path according to the recommended personalized learning data, the processed data and the sentiment data of the data set;

[0087] Learning the principles of planning decision-making technology in path planning module 32:

[0088] Principle ①: First, we need to clarify the learner's planning goals and needs, including the learning topic, difficulty, time and other requirements, and collect the recommended personalized learning data, processed data and emotional data of the data set;

[0089] Principle ②: Use decision trees to build a planning model based on the recommended personalized learning data, processed data, and sentiment data of the data set;

[0090] Principle ③: Establish a learning knowledge graph based on the planning model, and then plan a suitable learning path based on the established learning knowledge graph data.

[0091] Usage process:

[0092] The personalized recommendation module 31 recommends personalized learning based on the student knowledge gap data evaluated in the data evaluation module 13 and the behavioral model data in the model construction module 21, and transmits the recommended personalized learning data to the learning path planning module 32. The learning path planning module 32 plans the learning path based on the personalized learning data recommended in the personalized recommendation module 31, the data processed in the acquisition processing module 11, and the emotional data of the data set in the data analysis module 12.

[0093] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. Intelligent education system based on multi-algorithm fusion, characterized by: It includes a processing and analysis unit (1), an extraction and construction unit (2), and a personality planning unit (3); The processing and analysis unit (1) is used to obtain and process the student's relevant learning data and other historical data, and then analyze the processed data to determine the student's learning behavior patterns, difficult problem-solving ideas, and psychological emotions; The extraction and construction unit (2) is used to receive the data in the processing and analysis unit (1) and extract relevant data and construct a model; The personalized planning unit (3) is used to receive the data in the processing analysis unit (1) and extract the model data constructed in the construction unit (2) to perform personalized recommendation and learning path planning; The extraction and construction unit (2) receives the data in the processing and analysis unit (1), and extracts relevant data based on the learning difficulty problem data and the student learning rules analyzed in the processing and analysis unit (1), and then constructs a thought model and a behavior model based on the extracted relevant data, and transmits the constructed thought model and behavior model data to the personality planning unit (3). The personality planning unit (3) makes personalized thought recommendation and personalized learning recommendation based on the thought model and behavior model data constructed in the extraction and construction unit (2) and the data processed in the processing and analysis unit (1), and then plans the learning path based on the personalized learning recommendation data and the psychological data analyzed in the processing and analysis unit (1).

2. The multi-algorithm fusion intelligent education system according to claim 1, characterized in that: The processing and analysis unit (1) comprises an acquisition processing module (11), a data analysis module (12), and a data evaluation module (13); The acquisition processing module (11) is used to acquire student-related learning data and other historical data, wherein the acquired student-related learning data includes learning conditions, difficult questions raised by students and psychological emotional data, and then process the acquired student-related learning data; Learning data include students’ academic performance, learning behaviors, learning styles, and preferences; The data analysis module (12) is used to receive the data processed in the acquisition processing module (11), and use the clustering analysis algorithm to analyze the learning rules of students based on the processed learning situation data, and then use the knowledge solution algorithm to analyze the difficult problems raised by the processed students and the relevant knowledge points of the difficult problems in the database. The hierarchical clustering algorithm is used to perform psychological analysis based on the processed psychological and emotional data, and then the multi-algorithm fusion technology is used to fuse the data analyzed by the clustering analysis algorithm, the data analyzed by the knowledge solution algorithm and the data analyzed by the hierarchical clustering algorithm into a data set.

3. The multi-algorithm fusion intelligent education system according to claim 2 is characterized in that: The data evaluation module (13) is used to receive the student-related learning data processed in the acquisition processing module (11) and the student learning regularity data in the data set in the data analysis module (12), and use the shortcoming evaluation technology to evaluate the student's knowledge shortcomings based on the processed student-related learning data and the student learning regularity data in the data set.

4. The multi-algorithm fusion intelligent education system according to claim 3 is characterized by: The extraction construction unit (2) includes a model construction module (21) and a concept adaptation module (22); The model building module (21) is used to receive the student learning rule data and the analyzed difficult problem-solving ideas data in the data set of the data analysis module (12), extract the student's relevant learning rule data and difficult problem-solving ideas data, and use the training model technology to build a behavior model based on the extracted student learning rule data, and then use the training model technology to build a thinking model based on the extracted difficult problem-solving ideas data.

5. The multi-algorithm fusion intelligent education system according to claim 4 is characterized in that: The idea adaptation module (22) is used to receive the idea model data and the behavior model data constructed in the model construction module (21), and use the knowledge graph technology to perform knowledge reasoning of multiple problem-solving ideas based on the constructed idea model data, and then match the problem-solving behaviors in the reasoned multiple problem-solving idea data and the behavior model data to match a problem-solving idea with the highest probability of the student's problem-solving behavior, and at the same time recommend a problem-solving idea for the student to choose.

6. The multi-algorithm fusion intelligent education system according to claim 5 is characterized by: The personalized planning unit (3) comprises a personalized recommendation module (31) and a learning path planning module (32); The personalized recommendation module (31) is used to receive the idea model data in the model building module (21), the matching problem-solving idea data in the idea adaptation module (22) and the data processed in the acquisition processing module (11), and use the learning recommendation technology to make personalized learning recommendations based on the idea model data, the processed data and the matching problem-solving idea data, and the personalized recommendations include suitable learning resources, course content and exercise questions.

7. The multi-algorithm fusion intelligent education system according to claim 6 is characterized by: The learning path planning module (32) is used to receive the personalized learning data recommended in the personalized recommendation module (31), obtain the data processed in the processing module (11) and the sentiment data of the data set in the data analysis module (12), and use planning decision technology to plan the learning path based on the recommended personalized learning data, the processed data and the sentiment data of the data set.

8. The multi-algorithm fusion intelligent education system according to claim 7 is characterized in that: The planning decision technology implementation principle in the learning path planning module (32) is: Principle ①: First, we need to clarify the learner's planning goals and needs, including the learning topic, difficulty, time and other requirements, and collect the recommended personalized learning data, processed data and emotional data of the data set; Principle ②: Use decision trees to build a planning model based on the recommended personalized learning data, processed data, and sentiment data of the data set; Principle ③: Establish a learning knowledge graph based on the planning model, and then plan a suitable learning path based on the established learning knowledge graph data.

Images