摘要：Structured P2P network is highly efficient and low cost in resources search, but it only supports single-keyword precise search rather than multi-keyword fuzzy search. This paper puts forward a novel search algorithm FSA-MDSM based on semantic vector matrix, in which the P2P node resources can be transferred into a number of semantic vectors which are formed into a vector matrix in terms of a free semantic dictionary. And according to the similarity of vectors semantics, the vectors are classified into several sub-blocks, each of which is managed by a virtual node. Those virtual nodes with similar semantics will cluster into several special subnets by the way of exchange of the semantic information, and a semantics routing table will be generated. The table can guide a node to locate target resource rapidly. The simulation experiments demonstrate that by using FSA-MDSM algorithm, a fuzzy semantic search can be supported efficiently in the Structured P2P network and its original advantage can not be lost basically
