Content-based Search using Term Aggregation and Classification over Hybrid Peer-to-Peer Systems

In this paper, we propose a method to support contentbased search, one of the challenges in Peer-to-Peer file sharing system. The proposed method is based on a hybrid structure, which is a combination of a Chord ring and a balanced tree. The tree is used to aggregate and classify terms while the Chord ring is used to index terms of nodes in the tree. At every node in the tree, the system classifies terms as either important or unimportant. Important terms of a node, which can distinguish the node from its neighbor nodes, are indexed in the Chord ring while unimportant terms that are either popular or rare terms are aggregated to higher level nodes. Based on the classification, the system can process queries on the fly without the need of global knowledge. Therefore, our system can avoid the problem of bottleneck at nodes keeping global knowledge and the expensive cost of synchronization global knowledge among these nodes. We have done extensive experiments to validate the effectiveness and efficiency of our proposal.