The Research Of Structured P2P Network Model Based On Semantic Search

P2P (Peer-to-Peer) technology, as one of the important Internet technologies, has been more and more attention in computer industry in recent years. At present, although most of the P2P network application is supported by the P2P mainstream network model. But because its mainstream network model's own limitations of the algorithm is not suitable for some special needs of the P2P network applications, take semantic search as an example. So we need to find a more suitable network model.First, this paper studies and analyzes the principles of vector space model, and on this basis, puts up forward with an improved vector space model based on text clustering algorithm—ZH-Sem algorithm. This algorithm does not need more similarity between all clusters, has fast execution speed and high availability, and is applicable to large-scale file system. In addition, it does not require pre-determined the values of clusters k in the clustering process, this reduces the dependence of domain knowledge and improves flexibility and scalability.The paper then summarizes and analyzes the current major P2P network model and semantic model of development status and trends, and on the basis of the ZH-Sem algorithm, structures peer-to-peer model for further expansion, mentions a Chord-1evel P2P network model—Structured P2P Network Model based on Semantic Search (PNSS), and analyzes the information retrieval process and work process of the model. PNSS model designed to provide users with more efficient and accurate service functions, such as the release and deletion of information resources. Which enable the users to quickly and accurately find the necessary resources to improve the P2P network information retrieval precision and recall ratio.