Research On The Semantic Web Fuzzy Rough Ontology Driven Reasoner

Along with a large number of uncertain information as furture development of the Semantic Web, the existing ontologies have been inable to represent their real enough contents completely, to large extent, only some specific aspect of uncertain information can be solely deal with respectively. Fuzziness and roughness are not two opposite aspects, but take on a complementary relationship. In this thesis, fuzziness and roughness are combined and introduced into ontology to form fuzzy rough ontologies, which can perfect knowledge expression ability. Reasoning plays an important role in the development process of Semantic Web, and how can the uncertain information be represented and reasoned, an ontology inference engine that can deal with both fuzziness and roughness is needed.The existing inference engine are mostly description logic-based and rule-based, and their reasoning objects are precise ontology alone, and cannot be satisfied with the reasoning of uncertain knowledge in the real world. With the continuous development of intelligent technology, the reasoning requirements and thrust range continues to expand. At this stage, the existing inference engine is not suitable for fuzzy rough ontology reasoning. So, the thesis presents the semantic web fuzzy rough ontology model. According to the fuzzy rough ontology model, the thesis studies the fuzzy rough description logics, to lay the foundation for the description logic-based reasoning.This design of the driven reasoner is in order to meet the actual needs of customers; which can automatically select the appropriate reasoner to perform different reasoning tasks. The ways of reasoning include description logic-based reasoning and rule-based reasoning, if the customers’s requires are to test logic errors, then they can choose description logic-based reasoning. In terms of description logic-based reasoning, the Pellet and FuzzyDL reasoning ways can be chose. If the customers’s requires are mining implicit information, they can select rule-based reasoning. In the case of necessary, the user can combine the two kinds of reasoning sequencely. In this way, more needs for users themselves are considerd, it also can reduce the inference time and improve the efficiency of reasoning. Ultimately, driven reasoning by fuzzy rough ontology achieves though ontology reasoning and solves the reasoning problems of uncertain knowledge. According to the above methods, the thesis designs and implements a fuzzy rough ontology driven reasoner system FRODR. In order to verify the rationality and effectivity of the system, a test set is used to the experimental verification of the system. The experimental results show that the prototype system FRODR designed in the thesis can quickly and accurately reason fuzzy rough knowledge, and can be found that, the driven reasoner system was able to infer the fact contains information input detection logic errors, and lay a good foundation for the realization of intelligent information service.