Strum: A Semantic Trust-Based Miner Model for Detecting Influential Spreaders in Social Networks

Abstract

Social Networks have become embedded in every aspect of today people's life. The success of online social networks has a significant role in shaping trust and influence among participants. They are continuing to gain ground every day, opening the opportunity for several research movements. Great attention is now given for identifying influential spreaders in many application domains as recommendation systems, viral marketing, epidemic control and others. Most of the recent research focused only on the network structure with its geospatial information disregarding the semantics underlying these networks. In this paper, we propose a semantic trust-based miner model (STRUM) that efficiently detects influential spreaders in social networks for enhancing recommendation services. STRUM considers both geospatial information and semantic information in defining each user in the social network. Different experiments have been conducted to verify and validate the proposed algorithm against the latest published using two real location based social networks namely; Brightkite and Weeplaces are used in the experiments.