The Research Of Ontology Mapping Based On Computing Similarity In The Semantic Web

In the traditional areas, Semantic Web resources information or knowledge in a free or semi-structured text (HTML) to describe the data, and it is lack of clear specification of semantic descriptions. Semantic relationships between information resources that exist in an implied way are difficult to understand and dealt with by the computer. To solve this problem, the Semantic Web is coming. Since ontology is "shared conceptual model of the formal specification", and can be used to solve the resource sharing and reuse. Therefore, ontology can be realized as an important part in the Semantic Web.However, in the ontology construction process, due to the distributed nature of ontology and the creator has the knowledge and many different factors, heterogeneous ontologies are formed in the same field. To this end, we need to find the ontology mapping semantic relations between two ontologies to solve the problem of ontology heterogeneity.Ontology mapping is used to solve Heterogeneous ontologies in the most effective way, and to find and discover semantic relationship between different ontology, and express semantic relations in the formal way. Ontology mapping can be used to establish a corresponding relationship between ontologies, forming a pool of shared knowledge, and achieving interoperability between them, so as to solve the problem of heterogeneous ontologies. During the process of ontology mapping, the most important and the core content is the calculation of the similarity between ontology concepts, and if the similarity value is greater than a certain threshold to think, then there are some two ontologies mapping.This paper focuses on ontology mapping. First of all, the Semantic Web and ontology mapping methods were introduced. Then the development of semantic web and some of the status about ontology mapping were analyzed. In order to miss the defect of existing methods, we proposed a multi-strategy mapping that based on the element name, instance, and the structure.Finally, the ontology mapping prototype system NISS on multi-strategy has been designed and implemented. Then, the system has been tested by international organizations OAEI benchmarks standard test data set. Finally, the experimental data can be statistically analyzed based on recall, precision and F-measure and they are in the field of information retrieval. According to the results of the experiment, it is clear that the proposed method has good results.