Research On Ontology Of Uncertainty Based On Rule

With the rapid development of Internet, nowadays, the network has become increasingly difficult to meet the growing needs of people. Semantic web, as the next generation network, can solve the problems of intelligent information retrieval effectively and solve the defects and shortcomings of the existing network fundamentally, ontology is the key technology to support the Semantic Web. In addition, because of the uncertainty of the real world and the problems of heterogeneous ontologies which are caused by the difference of ontology developers, application purpose and application fields, research and explore the uncertainty ontology and its merging problems is very necessary. Therefore, in this paper, we introduced the uncertainty into the ontology, extended the traditional certain ontology and study on the uncertain ontology merging based on the research of uncertain ontology.In this paper, we proposed a model of uncertain ontology firstly, and introduced the fuzziness and incompleteness which are the common in the uncertainty knowledge into the model, so that the model can show the characteristics of the uncertain ontology in Semantic Web better, and it promote the ontology determining the transition to the direction of uncertainty. In the model of the uncertain ontology which is proposed in this paper, we chose the cloud model, fuzzy theory and rough set theory to represent the fuzzy concept and fuzzy-random knowledge, the fuzzy knowledge caused by the uncertain association of concepts and the incomplete knowledge, according to the advantages and disadvantages of different theories in the expression of common uncertain knowledge and interdisciplinary knowledge of these knowledge.In order to solve the heterogeneous problems of uncertain ontology, we studied the characteristics of uncertain ontology, analyzed its similarities and differences with the traditional deterministic ontology on the basis of the proposed uncertain ontology model, and then proposed a method of uncertain ontology merging based on rules. This measure divided the conflict problems which appear in merging of uncertain ontologies into four kinds:the conventional semantic conflict, structure conflict, uncertain semantic conflicts and the other conflicts, according to the features of different conflicts, the respective advantages of different theory on the representation of uncertain knowledge and the principle of merged ontology should maximization compatible the source ontologies, we constituted the combined rules and conflicts resolution rules, then merged the uncertain ontologies according to the rules, completed the ontology merging in the four levels of the concept level, attributes level, relations level and instances level, and solved the conflict problems at the same time.Finally, according to the above method, we designed and implemented the algorithm of uncertain ontology merging, and used the same domain ontology set which is provided by the OAEI2012and the uncertainty ontology set which is the transformation of OAEI2012test set, to validate the algorithm on two aspects:deterministic ontology merging and uncertain ontology merging. The experimental results show that, the proposed ontology merging algorithm can realize the uncertain ontology merging better, solve the heterogeneity problems, and has good expansibility.