A novel ranking strategy in hybrid peer-to-peer networks

Abstract

The state of the art of hybrid P2P-based information retrieval is still at its infant stage and confronted with many challenges. One of the most urgent problems is how to combine the retrieval results from different neighboring nodes into a single, integrated ranked list. In this paper, we propose a result merging algorithm to address the challenge. Our algorithm is deterministic and doesn't require neighboring nodes to provide any information for result merging, which makes it different from other algorithms that require cooperation from neighboring nodes. A variety of experiments demonstrate that the new approach is effective. © Springer-Verlag Berlin Heidelberg 2005.