Application Research of Data Mining Techniques in P2P Network

摘要

With the increasingly wide application of Peer-to-Peer, a large amount of data have been accumulated in Peer-to-Peer. Nowadays, in improving the search efficiency and system performance, researchers primarily focus on improving the P2P model, these studies have became bottleneckes in some extent. In this paper, data mining technology is applied to P2P network, and some further improvements based on SWLDRM are done, nodes are clustered based on the features of object stored by K-NN algorithm, applicable decentralized clustering algorithm is designed, new K-Chord modle was constructed, simulation experiments between K-Chord and SWLDRM are done and the result shown that K-Chord has lower maintenance cost and higher search efficiency than SWLDRM.