SenseTrust: A Sentiment Based Trust Model in Social Network

Abstract

Online social networks, as popular media and communications tools with their own extensive uses, play key roles in public opinion polls, politics, economy, and even governance. An important issue regarding these networks is the use of multiple sources of publishing or re-publishing news and propositions that can influence audiences depending on the level of trust in these sources between users. Therefore, estimating the level of trust in social networks between users can predict the extent of social networks’ impact on news and different publication and re-publication sources, and correspondingly provide effective strategies in news dissemination, advertisements, and other diverse contents for trustees. Therefore, trust is introduced and interpreted in the present study. A large portion of interactions in social networks is based on sending and receiving texts employing natural language processing techniques. A Hidden Markov Model (HMM) was designed via an efficient model, namely SenseTrust, to estimate the level of trust between users in social networks.