Epidemic Information Diffusion: A Simple Solution to Support Community-based Recommendations in P2P Overlays

Epidemic protocols proved to be very efficient solutions for supporting dynamic and complex information diffusion in highly distributed computing infrastructures, like P2P environments. They are useful bricks for building and maintaining virtual network topologies, in the form of overlay networks as well as to support pervasive diffusion of information when it is injected into the network. This paper proposes a simple architecture exploiting the features of epidemic approaches to foster a collaborative percolation of information between computing nodes belonging to the network aimed at building a system that groups similar users and spread useful information among them.