Bipartite Consensus for Global Trust in Social Network Services

As social network services (SNS) are gaining popularity, establishing trust relationships among users in SNS is necessary for user interaction and decision making. However, the evaluation of global trust values (i.e. reputation) of users suffers from the curse of opinion divergence in the network. In this paper, we consider the scenario of opinion divergence within social networks, and propose a method to reach different global trust values between groups of users with controversial opinions in the network. We model the global trust opinion formation in discrete-time dynamics and introduce bipartite consensus as the approach to establish global trust in such circumstances. Such approach works upon the property of structural balance, and can be extended to more general situations where eventual positivity applies. Via our approach, non-trivial global trust can be reached within the network, which can guide users´ social behaviors, and support many SNS-based applications.