Research And Application Of Semantic Web On Concept Based Retrieval

According to the traditional information retrieval pattern, the information lacks a uniform semantic description. It is hard for the user to find more relevant information and to realize the semantic share in information resource. How to assign machine understandable meaning to the information resource, and how to reach the semantic searching, are challenging tasks for information retrieval.Semantic Web is the next generation Web, which is advocated by Tim Berners-Lee who is the father of WWW. In Semantic Web, every resource has a unique identifier, and the relationships between resources can be easily processed by computers. Semantic Web is an extension of the current Web and makes computers and people work in cooperation.Having analyzed the principles, technologies, means, and features of the traditional information retrieval in the current Web, the paper proposes a framework of concept retrieval system based on Semantic Web. It also describes the design idea and retrieval process of the system in detail, and proves the validity of the framework in theory.After analyzing the key technologies of the concept retrieval system, the paper gives our effective solutions of realizing the concept retrieval system. These key technologies include ontology methodology, knowledge acquisition, semantic inference, concept extending, and so on.Based on the theory analysis, we have successfully developed the Concept Based Paper Search System (CBPSS). CBPSS selects science papers related to computer science as experiment objects, and builds related domain ontologies. Compared to the traditional key word matching retrieval system, the system can find the internal relationships among resources, and represent more relevant resources to the users. CBPSS resolve the problem of lacking semantic information in the traditional information retrieval system, and satisfy the request of the users for semantic retrieval.