DRank: Decentralized ranking mechanism for semantic community overlays

We propose a decentralized ranking algorithm for finding top-k users in a semantic social overlay based network. In large semantic networks the problem of finding top k users (or nodes) with respect to a particular topic is important. Be it a co-authorship graph where a author is looking for other top k authors with respect to a topic, or the problem to find top k influential nodes with respect to an interest (or topic) in a social network. In large networks, global knowledge is difficult to keep at individual nodes because the networks are (i) dynamic in nature and (ii) usually scale to very large numbers. Hence there is a necessity to design algorithms based on local neighborhood. Our proposed algorithm exploits social links and uses local information only. The algorithm scales upto any size of the network. The experimental results on both synthetic and real-world datasets show the effectiveness of our approach.