Research On P2P Information Retrieval With Semantic Support

P2P is one of the key technologies of restruction of Internet in the future. Its core idea is that all Peers involved have equal functions and responsibilities, can join or leave freely. They formates a truly dynamic network. As more and more resources are store to P2P peers, we need resources locating and retrieval capabilities. Thus, P2P information retrieval technology become one the research focus which is integration of P2P and information retrieval.The relevant theory and technique involved P2P information retrieval technology are compared and analyzed. As the existing system is difficult to provide semantic orgnization and it can not support the semantic resources locating and routing, on the basis of theory of resource space model, ontology, vector space model and CAN, it put forward A P2P Semantic Information Retrieval Model PSRM.PSRM is composed of unstructured layer and structure P2P layer and semantic retrival layer. In structure P2P layer, it optimized CAN that to adopt Polar coordinate and firstly uses the expansion space and power series algorithm, greatly reducing logic routing hops. Semantic retrival layer is composed of RSM_CAN and Ontology vector space model. The RSM_CAN origined of combining of the CAN and Resource Space Model. It organize and management of distributed resources, realizats the classification of resources. Ontology Vector Space comes from the traditional Vector Space Model theory, however documents vector is conversed to ontology concepts and semantic properties. Finally, the PSRM's simulation data analysis shows that it not only can support semantic information retrieval, but also effectively meet customers' needs.