Semantic enrichment for ontology mapping

System interoperability is an important issue, widely recognized in information technology intensive organizations and in the research community of information systems. The wide adoption of the World Wide Web to access and distribute information further stress the need for system interoperability. Initiatives like the SemanticWeb strive to allow software agents to locate and integrate data in a more intelligent way via the use of ontologies. The Semantic Web offers a compelling vision, yet it raises a number of research challenges. One of the key challenges is to compare and map different ontologies, which evidently appears in integration tasks. The main aim of the work is to introduce a method for finding semantic correspondences among component ontologies with the intention to support interoperability of Information Systems. The approach brings together techniques in modeling, computation linguistics, information retrieval and agent communication in order to provide a semi-automatic mapping method and a prototype mapping system that support the process of ontology mapping for the purpose of improving semantic interoperability in heterogeneous systems. The approach consists of two phases: enrichment phase and mapping phase. The enrichment phase is based on analysis of the extension information the ontologies have. The extension we make use of in this work is written documents that are associated with the concepts in the ontologies. The intuition is that given two to-be-compared ontologies, we construct representative feature vectors for each concept in the two ontologies. The documents are ”building material” for the construction process, as they reflect the common understanding of the domain. Outputs of the enrichment phase are ontologies with feature vector as enrichment structure. The mapping phase takes the enriched ontology and computes similarity pair wise for the element in the two ontologies. The calculation is based on the distance of the feature vectors. Further refinements are employed to re-rank the result via the use of WordNet. A number of filters, variables, heuristics can be tuned to include/exclude certain mapping correspondences. The approach has been implemented in a prototype system iMapper and has been evaluated through a controlled accuracy evaluation with a set of test users on two limited but real world cases. The system is tested under different configuration of variables to indicate the robustness of