Research Of Similarity Computing And Ontology Mapping Approaches

Ontology has been considered as an important theoretical basis in many areas such as information integration, semantic web and knowledge management. However, in many cases, for the same concept, there will be lots of different ontology defined by different areas. Even in the same field, a different agency will define a different ontology. Therefore, the heterogeneous problem between ontology will emerge. To solve the problem of heterogeneous between ontology, realize the interoperability between different ontology, and thereby achieve ontology-based knowledge sharing and integration, a bridge of semantic mapping must be set up between ontology. That is, ontology mapping, as commonly known. But now, most of the ontology mapping methods are for the application-specific ontology, and they are usually based on the characteristics of their own. So, they can not adapt to environmental changes, namely, a relatively common ontology mapping mechanism isneeded.This paper studies the ontology mapping methods, analyzes their characteristics and existed problems, and then improves a new computing method of concept similarity in ontology mapping to enhance the versatility, adaptability of ontology mapping, and the overall of the similarity computing method. Finally, on this basis, it describes an ontology mapping architecture based on the computing method.The similarity calculation between concepts plays an important role in ontology mapping. In this paper, information that affects the similarity calculation, such as the instances of the concept, the hierarchical structure relationship, and some attributes are fully considered. And it designs an integrated semantic similarity calculation model. That is, it computes the similarity of concepts from different views, and then integrates the three similarity weights to get integrated semantic similarity between concepts. By using the same semantic similarity calculation model, users can enable the model common to all the ontology by changing the three similarity weight values to adapt to the mapping between different characteristic ontology.Finally in this paper, two small ontology related students' information are constructed, and then a simulated experiment is carried out by using the integrated semantic similarity calculation model that has referred. Experimental results show that the model can be used among ontology mapping, and it can embody the semantic relationship between concepts more effectively, make ontology mapping method more comprehensive, more applicable, and more versatile.