SWRL Based Semantic Relevant Discovery And Its Application On Ontology Mapping And Integration

Ontology is built for knowledge sharing and reuse. However, the Semantic Web is constituted by a variety of information sources and each source expresses the knowledge by building its own ontology. As the ontologies are built at different time in different environment with different purpose and background, even for the same problem, due to the different builder with their own knowledge and understanding. It may be diverse a lot. These will cause conflict among these ontologies. So it is quite difficult to share the information among such sources, in order to solve the problem, semantic mapping and integration is put forward.Ontology mapping and integration are two key technic for the Semantic Web. Finding the semantic relationship among the ontology concepts is crucial for the mapping and integration. Currently, computing similarity and relevant is the main approach for the semantic association. By computing the similarity, it can help to infer which concepts among these ontologies are similar and which are not. However, no matter the similarity computing approach is instanced based or heuristic based, they can not discover some connotative relationship, even more, some relationship can not be got from the similarity computing approach. So it usually needs manual work, it has its limitation.This dissertation proposes a new approach (SWRL Based Semantic Relevant Discovery, SSRD in short). It concerns implementation with Protege OWL Plugin for SWRLJESS, based on the OWL concrete syntax of the SWRL proposal. It intends to help reasoning OWL ontologies with JESS combined with SWRL rules. By establishing rules and reasoning logically, SSRD can discover connotative relationship among the ontology concepts effectively. It makes a new attempt for the ontology mapping and integration approach and is a good complement for the similarity computing approach.SSRD refers to the family model from Christine Golbrelch, draw out some classification and relevant properties, re-establish some core concepts, instances and relationship, set up its own family ontology FO, which includes an OWL ontology representing the family usual relationships, and a SWRL rule base representing the dependencies between those relationships. SSRD take FO as an example to illustrate the process of finding concept relationship by SWRL rules and JESS, not only syntactically and semantically, but also inference to get all the inferences.