Using structured P2P overlay networks to build content sensitive communities

This paper details a proposed peer-to-peer system, which allows a user to join communities of other like-minded users in order to exchange files. Utilising the routing capabilities of Pastry, the proposed system includes an indexing service, which facilitates the creation of virtual rendezvous points for users with similar interests (manifested by shared keywords). Users are described by the content they store. By using vector space modelling techniques users can be grouped together to form content sensitive communities. The system is built to serve as the basis for a distributed archive of research papers. The system is designed to improve the efficiency of file searches over current P2P file sharing applications. Search requests result in a comprehensive set of relevant documents being returned as searching are based on semantic meaning rather than literal matching.