An Combination Approach to Tackling Semantic Conflicts Based on RDF Model

Within semantic Web environments, ontologies have been adopted as their conceptual backbone. An important task for Web information integration and sharing is to resolve semantic conflicts because of semantic heterogeneity and information incompleteness. But it is rather difficult for ontology engineers to integrate Web information including semantic heterogeneity and conflicts. In this paper, a combined approach is proposed for detecting and handling semantic conflicts from ontology information. Using rule reasoning based on RDF model theory, we can automatically check where semantic conflicts probably exist, and further locate so called conflict points of ontology information. Then based on Bayesian network method, semantic conflicts can be controlled and avoided by computing semantic overlap and its posterior probabilities of concept nodes from conflict points. Users can issue semantic based query by interaction interface and achieve information sharing with conflicts and uncertainty.