Research On The Construction And Application Of Knowledge Graph Oriented To Programming

At present,the knowledge graph has been widely used in various fields,and its application and achievements in the field of education are particularly significant.The knowledge graph is playing an increasingly important role in education and teaching.Programming learning is an important branch of the education field.For beginners,there are still many challenges,such as a long learning cycle,lack of effective learning aids,and difficulty in obtaining timely help in programming practice.These challenges bring great inconvenience to programming beginners,and are detrimental to their knowledge mastery and ability improvement.Based on the above content,based on the analysis of the characteristics of programming language disciplines,this article focuses on the semi-automatic construction of programming language knowledge graphs and the application based on graphs.The main work of the thesis includes three aspects:First,a method for artificially constructing programming language ontology is proposed.This method combines the characteristics of programming language disciplines and makes relevant improvements based on the "seven-step method".The main steps include core concepts and hierarchical relationship determination,attribute determination and instantiation,and experts Audit and coding.Secondly,for the problem of programming language ontology storage management,a graph database storage strategy based on ontology element division is proposed.This method divides ontology elements into structure layer and instance layer,and completes the ontology model to the graph database model based on the mapping rules of the corresponding elements.storage.The experimental results prove that the proposed method is feasible and effective.Secondly,based on the programming language ontology construction,a semi-automatic method is proposed to complete the expansion of the programming language knowledge graph.Among them,for the problem of multiple relations in information extraction,a multi-strategy fusion information extraction method is proposed.This method uses a rule matching method and clustering algorithm that integrates term relevance to complete directional relationship extraction,and uses dependency syntax analysis to complete non-directional relationship..Aiming at the problem of knowledge contradiction or redundancy,an entity alignment method based on the thesaurus is proposed.The method constructs the thesaurus to standardize the expression of the extracted knowledge entities,and combines similarity calculations to supplement similar entities for the thesaurus.Correct.The experimental results show that the proposed method effectively completes the enrichment and improvement of the programming language knowledge graph.Finally,the applied empirical research is carried out on the constructed programming language knowledge graph.A programming language graph evaluation method based on weighted co-occurrence analysis is proposed.The method firstly counts the co-occurrence elements of the exercise label set in the practical application question bank and the knowledge entity set in the programming language knowledge graph,and uses the knowledge points in the question bank as the amount of exercises.The weighting basis is used to calculate the weighted co-occurrence coverage rate of the programming language graph to the problem labels.The experimental results show that the completed programming language graph is complete,which verifies the effectiveness of the programming language knowledge graph construction method proposed in this paper.