Research On Ontology Building And Semantic Integration

Web has become the main platform for Internet information exchange. With the development of WWW and increasing data, it becomes very important that how to let computers understand the meaning of data and provide convenience for people's searching for data what they need indeed.Under this background, Tim Berners-Lee who invented Web advanced the conceive of Semantic Web in 1998.Semantic Web would be an extension of today's web, well-defined meaning would be put in messages, so computers could understand the messages, consequently the cooperation between people would become very well. Semantic Web would adopt multiple-layered framework to achieve the dream. Ontology layer lies between the turn of document depict and knowledge reasoning, so ontology building plays a key role in implementing semantic web.On the basis of the evolution of ontology describing language, we first investigate the methodology for ontology building, simultaneously combine the three basic methods for ontology constructing which includes manual construction, reusing existing ontologies and semi-automatic approach. Then we aim at some automatic methods of taking out semantics in structured data and semi-structured ones. Finally we propose that the generic method of ontology building will mainly depend on manual construction assistant with the semi-automatic methods.Application is the objective of ontology construction. The elementary principle in semantic integration is expatiated. On this basis, aiming at nested XML document, we design its semantic integration model on the premise of retaining document structure using XML as data source and RDF as ontology schema.The research of this paper is in hopes of making contribution to the development of Semantic Web.