Finding leaders from opinion networks

Abstract

This paper is motivated to utilize results from opinion mining to facilitate social network analysis. We introduce the concept of Opinion Networks and propose a PageRank-like algorithm, named OpinionRank, to rank the nodes in an opinion network. This proposed approach has been applied to real-world datasets and initial experiments indicate that the sentiment information is helpful for finding leaders of online communities and that the OpinionRank method outperforms benchmark methods that ignore sentiment information.