An Adaptive Search Method Based on the Interest in Gnutella-like Network

Yang Lei,Yang Hao,Wang Can,Qin Zhiguang
DOI: https://doi.org/10.1109/iccasm.2010.5623202
2010-01-01
Abstract:Search performance is the key issue for the development of the P2P network. Aiming at the flaw of existing search methods in unstructured P2P environment, an adaptive search method based on the interest is proposed. The interest overlay network is constructed according to the clustering characteristics of the peers. Due to diversity interest of every peer, the interest-domain of the peer is subdivided into several interest-themes, every peer joins in several interest groups at one time according to the similarity degree of the interest-themes. The interest relation is self-managed by the peer. Based on the similarity between the query and the interest-themes, the requesting peer adaptively guides the direction of the query. Simulation experiments show that the proposed method can effectively improve the search performance.
What problem does this paper attempt to address?