A Research On The Method Of Mapping Among Heterogeneous Ontologies

In recent years, ontology has been used broadly in the domain of knowledge share, merge and integration. But with the explosively increasing number of ontology, it is inevitable that a large amount of ontologies describing the same domain and similar concepts are developed independently and there may exist many ontologies for the same domain with leading to the heterogeneity problem.How to deal with the heterogeneity problem to realize the communication among ontologies. This thesis proposed a method which based on the Bayesian and the theory of Ontology Constructing in the view of ontology mapping to work on the heterogeneity problem.This thesis focuses on research of exploiting all kinds of ontology characteristics including concepts, attributes, instances, relations, context structure, descriptions and data type to realize the mapping between heterogeneous ontologies. The main content of the research includes: based on analyzing the key points and limitations of the current typical ontology mapping methods, proposed the model of HOMMC ( Heterogeneous Ontology Mapping base on Minimum Cost) which based on the Multiple Policies Similarity Computation and Bayesian. Analyzing the concrete steps of the mechanisms and corresponding algorithms description in detail and compared with some current methods, realizing the Heterogeneous Ontology Mapping base on Minimum Cost algorithms in java programming language and choosing experiment example to test the feasibility for proving that ontology mapping mechanism on HOMMC can effective complete the heterogeneous ontology mapping task.Proposed ontology mapping mechanism on HOMMC still needs the involvement of domain experts, which asks for automatic execution through further research and tells the complexity of ontology research.