Discovering interest-based community structure in Gnutella

Community structure is an important characteristic in complex networks. It has been discovered in many networks such as social and biological networks, which is significant to the application of these networks. However, very limited work has been done in the Study on the community structure in P2P networks. In this paper, the interest-based community structure of Gnutella (an important P2P network) is studied. We capture node information from Gnutella and propose a new method to measure the interest between nodes to construct interest-based network. A new. random walk based method is also designed to analyze community structure in interest-based network. Our results show that interest-based community structure exists in global Gnutella and local area of Gnutella, and the larger the area is, the stronger it will present.