Research On Ontology Mapping Algorithm Based On The Computation Of Concept Similarity

As the number of ontologies keeps growing, information exchanging and processing between different ontologies become an urgent problem. Ontology mapping is the most widely used and effective techniques in establishing relation between different ontologies. And concept similarity computation plays a crucial role in the process of effective ontology mapping. This paper focuses on the combination of different methods in concept similarity computation in mapping algorithm, proposing a mapping method which computes the concept similarity in the way that combines mixed and compound similarity measure.On the basis of studying on the theory and mapping method of ontology at home and abroad, we analyze the concept of the ontology of each name, attributes, instances and context structure. Then we give the methods of concept similarity computation based on concept's name, properties, instances and structure, proposing an improved ontology mapping discovery strategy. In the strategy, the above similarity computation methods are in mixed and compound way of combining to get the concept similarity between different ontologies. We have ontology mapping discovery done by getting the set of mapping concepts which is generated by the concept of similarity matrices. The mapping results are tested by computing the recall and precision.In this paper, we use the test set called benchmark provided by OAEI to test the performance of our mapping algorithm. We have our mapping results compare with the results generated by the mixed similarity computation based algorithm and the compound similarity computation based algorithm. It shows that the mapping algorithm proposed in this paper is valid. Also, we have our algorithm compared with three mapping system which participated the competitions held by OAEI, giving the corresponding analysis result.