| Learning is always an important need of people.In contemporary society,people are no longer satisfied with the traditional classroom learning in school.More and more learners begin to turn to the online learning mode on the high-speed Internet.More and more researchers also begin to pay attention to how to provide more "personalized" and "intelligent" services for learners.For the rapid development of this emerging field,knowledge graph has played a very important role,and its related research is of great significance for learners,teachers and researchers of personalized learning guidance system.This article takes the educational graph composed of teaching knowledge points and their connections as the research object,combined with complex network analysis methods,and designs and implements an analysis system that can satisfy all kinds of users’ multi-granularity and multi-angle queries.Firstly,it introduces the research and application status of online learning and knowledge graph.Through the summary of the results of the learning platform,personalized guidance,integrated environment and other aspects,combined with the research progress of knowledge graph,especially the application status of domain knowledge mapping,this paper expounds the necessity of the research on education graph,explains the necessity of educational graph research in this paper By analyzing the relationship between complex network theory and knowledge graph,the advantages and feasibility of applying complex network analysis method to education graph are illustrated.Then,through the requirement analysis,the user’s requirement objectives are defined,including different granularity query,multi-angle analysis and visual display.The basic function modules of the system are determined,including data management module,data retrieval module,network analysis module and graph visualization module,and detailed functional design and operation process design for each part.Among them,for the key needs of users,that is,the importance analysis of knowledge points,based on Page Rank algorithm,this paper analyzes the differences between the education graph and the traditional complex network,and puts forward an improved algorithm: combined with the topological statistical information of the graph,it puts forward a non-uniform initialization method;combined with the pre-relationship between knowledge points,a method of reverse transfer of node importance is proposed;combined with learners’ learning habits,this paper proposes a method to correct the medium granularity dilemma.Experiments verify that the algorithm in this paper can better reflect the importance of knowledge points under the premise of ensuring basic time efficiency.Finally,based on the overall design of the system,complete the code realization of each part of the function,integrate it into an educational graph analysis system,meet the goals set by the demand analysis,and pass detailed tests to ensure the stable operation of the system,provide services for all kinds of users,including learners. |