Research On The Application Of Description Logics In Intelligent Information Retrieval

In recent years, with the development of global networks, the information is growing at analarming rate every day. In such a period of information explosion, the traditional informationretrieval has been difficult to meet needs of the people. Along with the widespread application and therapid development of information retrieval technology, the intelligent information retrievaltechnology is an important developing direction in the future. The intelligent information retrieval hastwo notable features: it must construct a knowledge system based on some kind of semantic model,which is the premise and foundation of the intelligent information retrieval; then the resource objectshould be made the semantic annotation based on the knowledge system, to form the knowledge base.This thesis introduces the theory of description logics into the process of information retrieval.On the one hand, unstructured information is transformed into standardized knowledge representationwith the domain ontology using the semantic annotation technology. On the other hand, establishingthe mapping between domain ontology and description logics, the domain ontology has the capabilityof automatic reasoning, while the domain ontology can be optimized. Based on the above, reasoningservices are used in the information retrieval, and we can do information organization by domainontology and do knowledge reasoning by description logics, to realize the intelligent and automationof the unstructured information management and retrieval.Firstly, we give out the construction method of the domain ontology which is suitable forunstructured information management, and then analyze organization and storage methods ofsemantic annotation of the unstructured information. Secondly, establishing the mapping betweendomain ontology and description logics, retrieval demands and retrieval objects which descibe innatural language are transformed into the description logics representation.Analyzing the descriptionlogics reasoning algorithm，we select the inference engine which is suitable for the informationretrieval. Finally, we design the architecture of the intelligent information retrieval system based ondescription logics and several related core modules, and then give out the system prototype.Based on the above, this thesis designs corresponding business scenarios and apply this methodof intelligent information retrieval in this scenario to verify the feasibility and superiority of thismethod.