Research On Data Integration Based On Ontology Technology

With the development and application of database and network technology, massive data are stored in heterogeneous and distributed databases or files. Lots of enterprises and central office demand to integrate the distributed and heterogeneous data to realize data share. Then the heterogeneous data can be queried and analyzed by users as a unified view.Therefore,how to resolve the system heterogeneity, syntax heterogeneity and semantic heterogeneity between data. sources becomes a critical research subject.However,the tranditional integration methods have difficulties to resolve the semantic heterogeneity because they can not formally describe the semantics of data. Ontology, as a formal explicit specification of a shared conceptualization, is able to represnt the knowledge of specific domain effectively. It can be used as the universal semantic model to express the semantics of heterogenous data.This thesis is based on the jack-up drilling platform design information management system of construction technical research project background and put forward the subject, due to the complexity of Marine information platform business information source, application unit, large amount of information, have between data source distribution around in heterogeneous system,syntax and semantics. So in order to deal with the problem more effectively,this paper will use the ontology semantic layer integrated heterogeneous problem.First of all to face the problems in the current data integration are analyzed in detail,and summarized the current development status and application technology of data integration found their deficiencies, combined with the advantages and disadvantages of existing data integration technology, ontology mapping algorithm in the data integration as well as query transformations in query processing algorithm were put forward in this part of the research.This paper will first existing heterogeneous data integration technology is studied and summarized. Key research and put forward the improvement of comprehensive operation according to the similarity of the ontology mapping method. This method combines the concept of instance, the relationship between the name of the concept, concept, structure and properties of five aspects of operations, comprehensive operation mapping relation are obtained. In query transformation step, puts forward a kind of based on domain ontology query process, and transform algorithm. In search conversion process, is proposed based on domain ontology query conversion method, local ontology mapping query transformation method and the method of building SQL statements. And through the experiment to verify this algorithm's feasibility and accuracy.