Research On The Mapping Method And Similarity Calculation Of OWL-based Ontology

The semantic web technology is one of hot topics in the current Internet technical research. The semantic web is not a kind of brand new Web, but an extension of the present one, and it can express the information in a well-defined and formalized way, enhance the interoperability of heterogeneous Systems, facilitate the communion of the knowledge, and enable computers and people to work in cooperation. Ontology is the fourth layers of the Semantic Web structure, which is the crucial technology that makes Web have the semantic nature. 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. There are many criterias for building ontology. The criterion proposed by T.R.Gruber is most influential, and it includes five regulations: clarity and objectivity, coherence, completeness, maximal extendibility, minimal commitment. With the increasing number of ontologies, and as the creator and modeling method may be different, we must come to an agreement in order to implement information sharing, and that is to say, in order to achieve knowledge reuse and sharing among different ontologies, it is necessary to establish mapping among ontologies.The key of ontology mapping is the computation of similarities. After introducing the development of web research, this paper talks about some technologies related to semantic web, including the concept of ontology, the classification of ontology, the OWL description language of ontology. Then the paper classifies the mapping methods, and introduces computing methods of similarity in several mapping projects. After researching we find that the attribute and the relation of concept are very important to the concept of ontology, so that not only the similarity of concept name should be considered, but also the attribute and relation and the structure of concept should be taken into consideration.So, this paper focuses on the research of the concepts' hierarchy information and the relevant semantic mapping mechanisms in OWL, and then proposes an integrated ontology mapping approach ACAOM based on the attribute, hierarchy information, OWL definition of concepts. In this method, edit distance algorithm is used to compute the similarity between the concept name and concept attribute, the graphics method is combined to find the semantic neighbor of concept in certain semantic radius, the OWL semantic similarity of concept is worked out by the setting of the weight of OWL semantic relation together with edit distance algorithm, and finally four similarities are weighted to get the semantic similarity of two concepts. Considering all above, we import two ontologies to explaining the process of calculating the whole similarity and verified the method of calculating concept similarity based OWL ontology mapping.