Research And Implementation Of Information Retrieval Based On Semantic Expansion And Matching In P2P

Information retrieval is an important application in P2P. Problems such as unrelated results and low accuracy exist in the current keyword search based P2P system. With good conceptual layer architecture and logical reasoning support, Semantic Web technology can improve the computer's capacity of automatic analysis and intelligent processing during information searching. This paper studies the utilization of Semantic Web technology to improve the efficiency of information retrieval in P2P system.In this paper, the vector space model and query expansion technique in traditional information retrieval are studied, and a distance-based matching algorithm for semantic similarity is put forward on the basis of OWL ontology expressed resources and the improved distance-based semantic similarity calculation model. Furthermore, we analysis the features of information retrieval in P2P system, and propose a routing algorithm based on the semantic similarity of the nodes through applying the distance-based matching algorithm in P2P. The design idea of this algorithm is to express the node, document and request information in a vector form, perform synonyms and subclass expansions through ontology, and to calculate the semantic similarity between the vectors. We determine the matching success and query path forwarding in accordance with the semantic similarity values to achieve accurate matching of information on semantic level, and to make the query forwarding more directional in P2P.In this paper, a P2P information retrieval system based on semantic expansion and matching is designed, implemented and tested. The test results show that both the recall ratio and precision ratio of this system is better than those of keyword-based information retrieval system.