Font Size: a A A

Construction And Application Of Knowledge Graph-based Politic Repository

Posted on:2019-08-17Degree:MasterType:Thesis
Country:ChinaCandidate:X F XuFull Text:PDF
GTID:2428330569996101Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the application and development of the knowledge graph,the knowledge graph has brought new knowledge learning and knowledge retrieval to the basic education.The construction of political repository based on knowledge graph and its application were studied in this thesis.According to the comparative analysis of the status of knowledge graph and related methods at home and abroad,the modeling of political knowledge system and the acquisition of its knowledge base data were completed.On the question answering of political repository,the correct rate of the system has been improved by its optimal design.Combined with the needs of practical engineering,we designed and implemented a RDF repository data generation system based on Web.By using the web crawler,the automatic semantic annotation of encyclopedia resources was realized to generate RDF data,which made the political knowledge base more complete.The main work of this thesis includes:(1)Constructing a political repository on knowledge graph.In the construction of repository,we first use the artificial construction ontology to build the ontology of political subject by defining the domain scope,building knowledge classification system,knowledge concept system,defining object relationship,ontology modification and ontology audit.Then through the semantic annotation,the modeling of the political knowledge system was completed.Finally,the knowledge was stored,and the knowledge graph of the political subject based on the ontology was generated.(2)Designing and implementing the RDF repository data generation system based on Web.Based on the regular grammar matching and sensitive word filtering,an automatic semantic annotation method was used to achieve the RDF data acquisition of encyclopedia political discipline resources.The system was built with Python language.It has not only friendly interface and simple operation,but also is suitable for RDF data expansion in other fields of repository.(3)Optimizing the question answering system of political repository.Through regular grammar and common attributes of political science,we designed and edited the templates of political problems,and generated related basic templates and spare templates in the process of political knowledge question answering,the method of template matching method for political knowledge memory problems was given,and improved the accuracy of the question answering system.
Keywords/Search Tags:Knowledge graph, Political repository, Ontology, Semantic annotation, RDF, Knowledge question answering
PDF Full Text Request
Related items