Research And Implementation Of Concept Hierarchy Based Ontology Matching Strategy

Web information has been growing explosively. As a result, mass of information makes it difficult for users, to search and maintain information. One of the reasons is that the information is simply organized. Therefore the computer can not deal with it in perspective of knowledge with understanding of information content. In order to make information machine-understandable, Tim Berners-Lee et al proposed the concept of the Semantic Web, in which ontology provides an online common understanding on the information in interoperability between agents. Ontology is developed in distributed way by different organizations and individuals, with inevitable ontological heterogeneities. Ontology matching is one of technical means to cope with ontology heterogeneities. Most of the current ontology matching systems focus only on the match quality, rather than efficiency. So it costs long time to operate when the ontology matching systems and methods applied in the real world ontologies, especially the large-scale one. In this paper, entities of ontologies are divided into hierarchies through the hierarchical and structural information and information granularity. The entities are matched when they are of similar hierarchy, based on the concept that more similar information granularity of entities implies more probable alignment between them. Through the hierarchies, the number of comparisons is reduced and computation time is saved. Finally, several famous test ontologies, such as Russia, are used to conduct the experiment. The results suggest that Concept Hierarchy Based Ontology Matching Strategy has a good performance in matching ontologies and shows promising future.