Font Size: a A A

Research On The Construction And Application Of Knowledge Graph Of Enterprise Legal Person

Posted on:2019-08-17Degree:MasterType:Thesis
Country:ChinaCandidate:X Y YuFull Text:PDF
GTID:2428330566989962Subject:Software engineering
Abstract/Summary:PDF Full Text Request
In the era of big data,in order to efficiently use a large amount of Internet data,domestic and foreign Internet companies have built knowledge graph.In many areas such as intelligent question and answer,semantic recommendation,and public opinion analysis,knowledge graph play an important role with t heir strong knowledge organization capabilities and semantic processing capabilities.However,the current knowledge graph for the government's big data applications are very scarce,and many of the publicly-researched studies have emphasized on certain aspects of the knowledge graph construction process in isolation.In view of the above issues,aiming at the government and enterprise's corporate big data field,a method of knowledge graph construction based on multiple data sources is proposed.Based on this method,this paper describes the construction process of knowledge graph of enterprise legal person from the aspects of knowledge extraction,knowledge fusion and knowledge processing.First of all,based on some of the official corporate datasets,a crawler system that integrates Baidu's webpage data sources is constructed.By analyzing enterprise-related multi-source datasets,authors extract main classes,object attributes,and data attributes,then build an enterprise legal entity ontology to provide upper-level data patterns for knowledge graph.Secondly,domain ontology is introduced to solve semantic heterogeneity problem in multiple data sources integration.In this paper,we build Karma models to describe the semantic mapping between concepts and data in ontology and complete the rapid integration of multiple data sources.Based on build Karma models,authors publish a unified RDF data and R2 RML model.Thirdly,based on the unified RDF data,Jena reasoning engine is used to complete the mining of the implicit knowledge and the revision of knowledge,mainly including rule reasoning,upper and lower bit reasoning,missing class completion,and knowledge consistency detection.Fourth,aiming at the storage and visualization of knowledge graph.a RDF data storage scheme based on graph database Neo4 j is designed.The RDF2Neo4 j interpreter is used to parse and store the RDF triple data.Finally,based on the Spring Data framework,an enterprise legal person knowledge graph application system is developed to realize retrieval and application of knowledge graph.This paper takes the enterprise data set provided by the government and Baidu Encyclopedia as the main data sources,and elaborates the construction process of the enterprise legal person knowledge graph from the perspective of multi-source heterogeneous data integration,and realizes the visualization system based on the knowledge graph.Through a feature-rich application interface,users and developers can use and maintain knowledge graph easily and transparently,which is of great significance to the application and disclosure of basic data in government departments.
Keywords/Search Tags:Knowledge graph, Data integration, Karma, Ontology
PDF Full Text Request
Related items