Research On The Semantic Web Rough Ontology Aided Knowledge Reasoning

The Semantic Web is an extension of the World Wide Web. The objective of the Semantic Web is, through the development of a series of computer-readable and processed language and technology in the expression of semantic information, to let both the automated processing of information method and the understanding of information machine much easier. Ontology-based knowledge reasoning is the key technologies of the Semantic Web;it is the basis for other intelligent operation to achieve in the Semantic Web. Currently, the researches of Semantic Web are targeted at the precise ontology. But the objective things cognition to people is often imprecision based. So there is a large number of vague and uncertain information needs to be expressed. Roughness is an aspect of the imprecision. Owing to the advantage, which it doesn’t need any other external information except for the data, the representation of imprecise knowledge can become better and easier by introducing it into the ontology. Knowledge reasoning system that is able to handle the rough ontology is needed to complete the imprecise knowledge of the reasoning.Most of the current knowledge reasoning systems are based on description logic and rule separately, and are developed for the precise ontology which can not be applied directly to the rough ontology knowledge reasoning. There is no suitable for rough ontology inference engine at present. Therefore, this thesis first proposes a Semantic Web rough ontology model. Then we study the rough set, rough description logics, rough rules, and knowledge reasoning for the rough ontology model. By combining description logic-based knowledge reasoning method with rule-based knowledge reasoning method, a rough ontology aided knowledge reasoning framework is proposed on this basis.Based on rough description logic knowledge reasoning method, firstly, it’s rough description logic reduction for the classic description logic; secondly, it’s processed by consistency checking and inclusion checking; finally, no logic error in the knowledge base is submitted to the top-level inference engine to avoid absurd to push the absurd. The method, the rough ontology aided rule knowledge reasoning, is through a combination of the underlying knowledge base submitted to construct precise rules and rough rules and then to find implicit knowledge. Rule matching, rough matching and conflict resolution are needed to be solved. To some extent, this method reduces the workload of the former one. Ultimately, knowledge reasoning supported by the rough ontology achieves rough ontology reasoning and solves the knowledge reasoning problem for the uncertain information.According to the above methods, this thesis designs and implements a rough ontology aided knowledge reasoning system ROKRS. Finally, a test set is used to the experimental verification of the system. The experimental results show that the design prototype system in this thesis is capable of rough ontology for effective knowledge reasoning, and its accuracy and efficiency are quite high.