Research Of Multi-Source Heterogeneous Data Fusion Based On Domain Ontology

With the rapid development of information technology,competition between oil companies has turned to the application of information.There are many independent application systems developed in different historical periods For example,the professional application software,database system and professional data warehouse were established in different platforms,and produced a large number of heterogeneous data.These data sources are independent and have different standards of data.Based on the above reasons,research on data fusion and data sharing can greatly improve the value of these data and.In order to solve these problems,this thesis introduces the domain ontology theory into the data fusion and proposed a data fusion framework based on ontology,which realizes the integration of heterogeneous data fusion in semantic by using a two-layer-ontology structure.This structure can provide complete and effective information service to users.In this paper we propose a data fusion framework based on the petroleum domain ontology,and discussed the key technologies.First,we proposed a semantic feature extraction method of heterogeneous data.The concept and relations of petroleum domain ontology are used to extract semantic features of data from different resources and improve the efficiency of feature extraction.Secondly,the mapping method of ontology elements is proposed.And the similarity between ontology elements is defined as the combination of four parts: concept similarity,attribute similarity,relationship similarity and instance similarity.Finally,a series of fusion rules are defined for the query result,and the fusion results are normalized.Finally,this paper carries out the practice of ontology based information fusion service,designs and implements an ontology based data fusion system,and verifies the feasibility of the mechanism.