Research On Formal Lattic Based Ontology Merging Method And Technology

With the proposal and development of the Semantic Web, ontology has been widely used. But because of the different knowledge structure, naming habits, design preferences of different experts, and distributed heterogeneous of ontology, ontologies developed by different experts while modeling on the same area have significant differences, which is an severe impediment to interoperability of ontology. It is known that ontology merging is an effective approach to resolve this problem.Currently, there are a lot of researches on ontology merging approach, and some systems have been implemented. But all of these methods have some problems:(1) Domain experts are needed during ontology merging, it is very time-consuming and laborious. (2) the great human impact. For the same field, different domain experts may use different viewpoints to merge ontologies, leading to inconsistent results. So this paper proposes that using concept lattice to merge ontologies in order to solve these problems.Concept lattice, also known as formal concept analysis, is a branch of applied mathematics. It including the extension and intension, can express semantic information understandable by computer more easily. Firstly, uses two-dimensional table to express formal context of the field, then extracts all tacit concepts and the relationship between concepts to cluster concept, in the last, generates concept lattice. The biggest advantage is that the tacit concepts and the relationship between concepts can be obtained from the data in the case of without the impact from subjective factors of developers.This paper uses the theory of concept lattice for overall description of the ontologies. First of all, translates ontologies to formal contexts(data storage fashion of concept lattice) directly, and then juxtaposes these formal contexts using rigorous and comprehensive mathematical principle of concept lattice, and finally generates a new ontology. Experimental result shows that, the ontology merging system designed in this papper realizes fully automatic ontology merging and has high accuracy.