Junior High School Online Learning System Based On Knowledge Graph

The development of knowledge graph technology has created rich and favorable conditions for mining,analyzing,constructing and drawing the relationship between multiple entities in the field of modern education.This paper utlizes knowledge graph technology,natural language processing technology and crawler technology to mine the relationship between student entity,knowledge point entity and topic entity,constructs the knowledge graph of junior high school mathematics and student,and designs and implements the junior high school mathematics wisdom online learning system.which completes the student-oriented automatic question and answer and topic selection function,assisting teachers in teaching.The specific research is as follows:(1)The construction of the knowledge graph in junior high school mathematics topics.First,utlizing the crawler tool crawls a certain amount of topic data from the web site,and completes the construction of the word vector through natural language processing techniques such as word segmentation,word2 vec and etc.The average highest similarity and keyword similarity calculation methods are used to calculate the stem and answer respectively.The calculation results are used as the variables of the multiple linear regression model to identify similar topics and complete the topic deduplication.Then,the relationship between the topic entity,the knowledge point entity,the entity and the entity is extracted from the topic corpus.Finally,the concept layer is designed to complete the solid filling,and the constructed knowledge graph of the topic is visualized.(2)The construction of automatic question answering model and automatic topic selection model based on topic knowledge graph.The automatic question-and-answer model uses a template-based matching method to add a manually named named entity through the general-purpose named entity toolkit provided by the word segmentation tool HanLP to complete the abstraction of the sentence.Second,the automatic question-and-answer model create a question template based on the abstract sentence and add the template number.Then,a custom dictionary is introduced as a corpus,and the Bayesian classifier is used to train,finally the natural language is converted into a body instance template,which the result is queried in the topic knowledge graph and returned to the user.The automatic topic selection model is based on the strategy of learning progress and learning level,and recommends suitable exercises for students.(3)Development of an online learning system.Based on the analysis of application requirements,user requirements and functional requirements,this paper designs and develops a B/S-based junior high school mathematics online learning system.The system implements the knowledge point learning function based on the automatic question answering model and the practice recommendation function based on the automatic question selection model to help students quickly learn knowledge points and practice personalized topics.At the same time,the system also provides functions such as viewing rankings,which is convenient for teachers to view student rankings,and which is beneficial for teachers to timely evaluate students' learning status and timely guide confusion.The system has been put into the actual teaching application scene of a junior high school.The actual operation situation in the past five months indicates that the online learning system of junior high school mathematics based on knowledge graph can promote the learning efficiency and enthusiasm of students.