Skip to Main Content
Finding related resources is important for assisting resource retrieval and recommendation in the P2P network. This paper proposes a method for discovering such clusters of related resources, which are called resource communities. Graph mining approach is applicable here. Discovering communities is based on bipartite graph which is composed of resources and keywords. The data is extracted from the analysis of users' search and download behavior. Experiment shows that this method is effective in discovering resource communities and improving the efficiency of resource retrieval.