Research On Automatic Evaluation Method Of Math Teaching Plan In Junior High School Based On Knowledge Graph

With the rapid development of education informatization,software and hardware products such as Internet of Things,cloud computing,artificial intelligence and other technologies have been widely used in basic education teaching scenarios.However,the informatization degree of teaching preparation is lagging behind,especially teaching plan evaluation and modification,which is the core of lesson preparation,has been stuck in the inefficient stage of manual evaluation by senior teachers.These technique cannot meet the needs of teachers to evaluate their own teaching plan design and optimize lesson structure.Decided by the peculiarity of the function of the teaching plan,the conventional text evaluation methods based on language analysis such as syntax analysis and rhetoric use are not applicable to the evaluation of the teaching plan texts.Affected by the characteristics of the subject,the teaching of mathematics in junior high school is an orderly explanation of the mathematics knowledge network.The evaluation of the teaching plans focuses on the coverage,prominence,and conversion stability of mathematics knowledge points.Therefore,it is of great research value to introduce the knowledge graph into the automatic evaluation of mathematics teaching plans,map the texts of the teaching plans to the space of knowledge maps,grasp the key indicators of the assessment task,and reduce the complexity of the assessment.This article mainly builds a junior high school mathematics knowledge map based on textbooks.With the help of the graph database Neo4j,it improves the rendering effect and query performance of the map.At the same time,it proposes an accurate knowledge point extraction method,which opens up the mapping between teaching plans and knowledge maps.The improved path matrix and graph algorithm realize the automatic evaluation of junior high school mathematics teaching plans.The specific research content is as follows:(1)Construct a junior high school mathematics knowledge map.Firstly,by analyzing the requirements for modeling the knowledge graph,and combining the structural characteristics of the mathematical knowledge network,construction strategies of the knowledge graph are defined.Afterwards,in order to meet the needs of high-performance graph query and graph operation in this experiment,graph database Neo4j is used to store junior high school mathematics knowledge atlas.Finally,the construction of the test knowledge graph is demonstrated through the instance query of nodes and relationships.(2)Methods of extracting knowledge points of teaching plan in subjects.Extraction of knowledge point in teaching plans is the core procedure that maps teaching plans text to knowledge map.First,the evaluation criteria of knowledge point extraction task are defined.The shortcomings of traditional keyword extraction algorithms TextRank and TFIDF in knowledge point extraction task are tested from the principles and sources.Then,analyze the characteristics of the two algorithms on this task from code level,and propose a TFIDF knowledge point extraction method based on word length,part of speech,positional relationship and dictionary correction.Compared with the traditional extraction methods,the improved method has significantly improved accuracy,which can support the task of mapping the teaching plan text to the knowledge graph.(3)Teaching plan evaluation method.According to the function of the teaching plan and the research purpose of the teaching plan evaluation in this paper,it is determined that the teaching plan evaluation consists of the teaching plan evaluation and characteristic knowledge points evaluation of the teaching plan.According to the distribution rule of the teaching plan map,the teaching plan is divided into three types,and the path matrix and the centrality algorithm are introduced as the indicators of the teaching plan evaluation.According to the physical meaning of the path matrix and the characteristics of the knowledge point sequence of the teaching plan,a method for judging the type of teaching plans was proposed.Through improving the centrality algorithm,the evaluation obtained the difficulty of the teaching plan;based on knowledge graph,association query and path matrix analysis obtained the supplementary knowledge point,abnormal knowledge point and knowledge point span evaluation of the teaching plan.Finally,the feasibility and practicability of the teaching plan evaluation based on knowledge graph are verified by examples.