Top-k Query Evaluation for Schema-Based Peer-to-Peer Networks

Increasing the number of peers in a peer-to-peer network usually increases the number of answers to a given query as well. While having more answers is nice in principle, users are not interested in arbitrarily large and unordered answer sets, but rather in a small set of ”best” answers. Inspired by the success of ranking algorithms in Web search engine and top-k query evaluation algorithms in databases, we propose a decentralized top-k query evaluation algorithm for peer-to-peer networks which makes use of local rankings, rank merging and optimized routing based on peer ranks, and minimizes both answer set size and network traffic among peers. As our algorithm is based on dynamically collected query statistics only, no continuous index update processes are necessary, allowing it to scale easily to large numbers of peers.