Title

Discovering Influential Nodes from Trust Network

Document Type

Conference Paper

Publication Date

2013

Publication Title

Proceedings of the 28th Annual ACM Symposium on Applied Computing

First Page

121

Last Page

128

DOI

10.1145/2480362.2480389

Abstract

The goal of viral marketing is that, by the virtue of mouth to mouth word spread, a small set of influential customers can influence more customers. Influence maximization (IM) task is used to discover such influential nodes (or customers) from a social network. Existing algorithms for IM adopt Greedy and Lazy forward optimization approaches which assume only positive influence among users and availability of influence probability, the probability that a user is influenced by another. In this work, we propose the T-GT model, which considers both positive (trust) and negative (distrust) influences in social trust networks. We first compute positive and negative influences by mining frequent patterns of actions performed by users. Then, a local search based algorithm called mineSeedLS for node add, exchange and delete operations, is proposed to discover influential nodes from trust networks. Experimental results shows that our approach outperforms Greedy based approach by about 35%.