The Study On Personalized Push Of Learning Resources Based On Knowledge Graph

With the innovation of modern education technology and the development of Internet technology,it is trend to push personalized learning resources for learners.The effect of personalized education on students' development is beyond any doubt.More and more personalized push methods and personalized learning supporting systems are proposed and developed.But most of the current personalized push methods are based on online learners' online learning behavior data to push,paying more attention to the behavior analysis and basic information data of learners.These methods connot consider the relationship between konwledge points based on cognition,which may lead the result of personalized push is too dependent on learners and others' online learing behaviors in the online learning platform.Based on this,this research carried out research on the push of personalized learning resources based on knowledge graph at the cognitive level.The main researches include:(1)Taking algebra in junior middle school as an example,proposing the construction method of knoledge graph of junior algebra,which provides the basis for the realization of personalized resources push based on knowledge graph.The specific process includes: Combining the logic and structure of junior middle school mathematics knowledge points,determining the process of knowledge graph construction;Determing the rules of constraint;Collecting the structured and unstructured data;Defining the Class,Individual and Properties of the knowledge graph of junior algebra;Experts verification of knowledge graph.This knowledge graph contains resources of knowledge points,similar exercises,types of questions and instructional videos and so on.Each exercise contains different knowledge points.And,in the later similarity calculation,we identify the knowledge points involved in this question,using these knowledge points to return knowledge graph to search and match resources in the knowledge graph.(2)Completing the automatic matching of algebra exercises and knowledge graph and similarity calculation.The personalized push of exercises and matching of knowledge points and other resources in the knowledge graph are all need completing similarity calculation of exercises and identification of knowledge points.This research takes the junior algebra as point of penetration,using the method of semantic sentence patterns,basing on the knowledge graph,identifying the knowledge points of exercises and types of questions of junior algebra automatically.And using the Cosine similarity to judge similarity of exercises further.By this method,the experimental results show that the accuracy rate of the identifying of knowledge points is as high as 96.77%,the accuracy rate of the identifying of types of questions achieved 73.33% and the accuracy rate of the similarity calculation of exercises achieved 75.61%.Finally,this research takes algebra exercises in junior middle school as an example,proposing the method of personalized push based on knowledge graph which includes push of similar exercises,push of knowledge points of exercises,push of types of questions,and push of relative instructional videos and so on.And clearing the strategies of personalized push of exercises resources of junior algebra.