OPTIMISING ONTOLOGY INTEGRATION THROUGH INTERMEDIATE ONTOLOGIES

The semantic web is increasingly being seen as a solution to manage knowledge content among heterogeneous and distributed information on the internet. Evolution of the semantic web is linked to a great extent to the evolution of various domain ontologies. It is necessary to formally define the mapping between ontologies to enable interoperability between applications in heterogeneous distributed information systems. The authors, in this paper illustrate how the fundamental problem of mapping between the global ontology and the local ontologies can be addressed, primarily through a newly developed WeGO algorithm. A mapping system for OWL-DL ontologies, where mappings are expressed as correspondences between conjunctive queries over ontologies, forms the core of this research work. The algorithm finds the semantically equivalent terms in local ontologies and uses them to build an intermediate ontology. The intermediate ontologies form the building block for a global ontology that will encompass the salient elements of the various local ontologies. It is further shown how the mapping system proves effective for the task of ontology integration through illustrative queries. Experimental data show that the query results obtained from the local ontology and global ontology match the results obtained from the intermediate ontology.