Research On Similarity-Based Ontology Mapping Approach

With the rapid development of the Semantic Web, the problem with Internet resources which confusion, repetition and redundancy has been greatly improved. As the core of the Semantic Web, the Ontology’s study has made a great break through, and has been applied in many research fields. However, makes the phenomenon appeared with heterogeneous ontology in some field, ontology heterogeneity hinders the interaction and resource information sharing, and become the bottleneck of the ontology development. Ontology mapping is the most effective way to solve the problem of the recognized. This technique is found and computing similar information of the concept in heterogeneous ontology according to certain strategy, then map these similar concepts, eliminate the ambiguity between ontology concepts, to realize the interaction and sharing of information resources. But at the same time, in terms of efficiency and automation of processing still exist many difficulties and problems.This paper introduces the research status and the research background and significance of ontology mapping technology at home and abroad firstly, elaborated the domestic and foreign comparison of classical ontology mapping tools and methods, And the respective advantages and disadvantages are briefly analyzed, detailed introduced the related theory and knowledge of ontology mapping involved, such as the semantic web architecture, the concept of ontology heterogeneous phenomenon.The two core parts were researched which on the ontology mapping in pretreatment and the comprehensive concept similarity calculation, two big bottleneck have been improvement and solution by separately use the minimum spanning tree of a graph clustering algorithm and rough set method of conditional information entropy, which are mapping problem of pretreatment and similarity fusion over reliance on expert experience, unable to adapt to automatic mapping. On which the algorithm is implemented, improvement, and the set of test cases used in benchmarks data, to verify the validity of the algorithm.Using the fusion algorithm for concept similarity is proposed in this paper, a ontology mapping system is designed. The system uses the method of graphical interface to reflects the ontology mapping process and the algorithm’s final effect.