Heterogeneous Data Integration Framework For Enterprise-class Research

Along with the development and popularization of network and distribution application, a large amount of heterogeneous data sources have been spawned; the need of solving the integration problem of these heterogeneous data sources was becoming more important. To resolve this problem, the concept of ontology is introduced in this paper. Utilization of ontology did well in solving the problem of semantic heterogeneity of data integration; however the other technologies of data integration could hardly achieve it.Looking back on the current research situation, since the semantic integration has many difficult points in the integration of heterogeneous data, the methods of ontology integration are put forward. The method is to resolve the semantic problem by describing database with ontology. Since there are three kinds of different application situations, three corresponding methods of data integration are put forward in the paper.The main work of the paper includes the aspects below:Researches on foundation: Firstly, the basic steps of data integration are offered. Then a system framework of data integration of four-layer mode is put forward. This framework uses RDQL language and conjunctive query as the tool of data query and spreads the usage range of conjunctive query, which can also be applied on the query of ontology.Global ontology based data integration: Built a global ontology for all datasources and a local ontology for every datasource, at the same time set up the mapping between ontologies and datasources, and then decomposed the global query statement based on the result of reasoning ontology, producing the query statements of different datasources to realize the goal of unified visit.Ontology-based Data Mining on Heterogeneous Database Environment: The integration of heterogeneous data: it mainly studies on designing the framework of data mining basing on ontology in heterogeneous database, and remark on the process of establishment of global ontology and topic ontology, as well as the process of establishment of database basing on the former one.Local ontology based data integration: An ontology-based approach to the problem of data interoperability is described, specifically focusing on the issue of query processing in a distributed heterogeneous setting. In particular, we discuss the mechanisms for metadata representation, distributed ontology mapping, and query answering.