A Document Recommendation System Based on Clustering P2P Networks

: This paper presents a document recommendation system based on clustering peer-to-peer networks. It's an unstructured P2P system. In this system each agent-peer can learn user's interest, then it helps user share and recommend documents with the other users. Since each peer in our P2P networks is a node, in order to cluster them, we import the concept of Group. Each group is composed of peers. The types of documents, which belong to a same group, are uniform. This paper presents how these peers help users to share and to recommend documents, and how they cluster into groups. Our experiment results show the advantages of the document recommendation system. Key words: Recommendation System Clustering P2P Reputation Management. 1 Internet (Personalized Recommend) Amazon.com CD Now.com MovieFinder.com 

• Type: Conference
• Year: 2007
• Where: CDVE
• Authors: Feng Guo, Shaozi Li
• Keywords: CDVE 2007, Document Recommendation, Recommendation System, Unstructured P2p System, Visualization