A Game Theoretic Approach for Modeling Privacy Settings of an Online Social Network

Users of online social networks often adjust their privacy settings to control how much information on their profiles is accessible to other users of the networks. While a variety of factors have been shown to affect the privacy strategies of these users, very little work has been done in analyzing how these factors influence each other and collectively contribute towards the users’ privacy strategies. In this paper, we analyze the influence of attribute importance, benefit, risk and network topology on the users’ attribute disclosure behavior by introducing a weighted evolutionary game model. Results show that: irrespective of risk, users are more likely to reveal their most important attributes than their least important attributes; when the users’ range of influence is increased, the risk factor plays a smaller role in attribute disclosure; the network topology exhibits a considerable effect on the privacy in an environment with risk. Received on 14 February 2013; accepted on 23 March 2014; published on 27 May 2014