A Study Of Ontology Mapping

Ontology is a philosophical concept originally, and it describes the natures of objects. In 1980's, scientific researchers introduced ontology into the field of artificial intelligence (AT), so ontology has new meanings. In the fields of computer science, ontology is a formal specification of a shared conceptualization. The main purpose of applying ontology is knowledge sharing and reusing. Researchers propose many criteria for building ontology. The criterion proposed by T.R Gruber is one of the most influential. It includes five regulations: clarity and objectivity, coherence, completeness, maximal extendibility, minimal commitment. Because there is not a common criterion for building ontology, so great deals of different ontologies are constructed in the same domain. These ontologies are heterogeneous. Ontology mapping can solve the problem of ontological heterogeneity.The key of ontology mapping is the computation of similarities. By far there are two usual methods of computing similarities. One is based on instances, and the other uses heuristic rules. But they have shortcoming respectively. In this dissertation, an improved method of computing similarities is put forward. It combines and improves the above two methods. Ontology is usually viewed as the sets of concepts, attributes and relations. Because attributes and relations of concepts play important descriptive roles, ontology mapping considers not only the similarities of simple concept pairs, but also the influence of attributes and relations to the concept similarity.In this dissertation, the similarity is computed with two aspects: first computing the semantic similarity, and then computing the descriptive similarity. Semantic similarity describes the degrees of similarity between two concepts. Descriptive similarity explains the similarity of concepts from the view of attributes and relations. It is the similarity of attributes or relations. And an ontology-mapping framework (Multi-way ontology mapping framework, MOMF) is designed. It is a kind of hybrid architecture. The framework has five components: application ontology, reference layer, shared domain vocabulary, local ontology and mapping model. In application ontology, there is an existed top-level ontology. This dissertation uses Wordnet. Reference layer offers a matching table of data types and other relative information. The shared domain vocabulary is a global term vocabulary of the shared domain. It is...