Intelligent English Grammar Exercises System Based On Knowledge Graph

As the most widely used language in the world,English is the key subject of basic education in China.How to help learners absorb English knowledge effectively has always been the common goal of English educators.With the development of 2.0-version educational informatization,more and more English learners tend to take English exercises online.Since grammar is the foundation for English,grammatical exercises are usually regarded as a top priority among many kinds of English exercises.However,quantities of grammar exercises have not been intelligently processed in the current online exercising platforms,and most of them still rely on manual analysis by experienced teachers.In this paper,knowledge graph technology is applied to English grammar learning,and an intelligent exercises system is constructed and implemented based on knowledge graph.The main research results are as follows:(1)Construct the knowledge graph of English grammar.We analyze the knowledge of English grammar and extract the related names,concepts,examples and structures;Then,we design the conceptual model of grammar knowledge graph,and utilize Neo4 j to store and visualize it.The English grammar knowledge graph consists of 18 semantic relationships and 1103 grammar knowledge entities.(2)Analyze the system requirement and design the corresponding functions of intelligent English grammar exercises system.According to the learning requirement of users,we design the functions of exercise classification,knowledge graph view,exercise recommendation,intelligent Q & A,and realize the system architecture design and database design.(3)Design and implement the core functions such as exercise classification,exercise recommendation,intelligent Q & A based on knowledge graph of English grammar.For exercise classification,we adopt SVM algorithm with knowledge graph feature,correct answer part of term feature and TF-IDF feature as the exercise feature vector;for exercise recommendation,we search knowledge graph and get the confusing knowledge points related to the user’s history mistakes;for intelligent Q & A,we classify the users’ natural language questions by SVM classifier,and use cypher query language to match answers from knowledge graph.As an initial attempt and application of the knowledge graph in English grammar,the research provides users with a relatively comprehensive and accurate English grammar knowledge network.At the same time,it matches the current developing direction of education intelligence.