Controlling privacy with trust-aware link prediction in online social networks

Abstract

In real life, the strength of the relationship among people indicates the level of trust. People tend to trust their friends, family members' opinions rather than unfamiliar one. Since this is a human characteristic and the phenomena of social network services which allow users to post their information, diaries, albums, and experiences, the need of identifying reliable people within social networks become a fundamental goal to prevent unreliable users accessing these rich information. In this study, we propose a capacity-based algorithm that adapts Advogato trust metric for identifying people of trust based on weighted relationships. Our method finds all possible trustworthy users who are connected to a target user in his/her personal network and prevents unreliable users from impacting on a personal network. We present experimental results that show how our algorithm performs better than existing alternatives.