Analyzing the Robustness of a Comprehensive Trust-Based Model for Online Social Networks Against Privacy Attacks

Security and privacy have been major concerns of Online Social Networks (OSN). Individual users as well as organizations utilize OSNs, such as Facebook, Twitter, and LinkedIn, to share information with other users within their networks. While sharing information, users are not always aware of the fact that an innocent action on their post by a direct friend such as a comment or a share may turn the post transparent to someone outside their network. In previous work we have devised a comprehensive Trust-based model that combines Role based Access Control for the direct circle of friends and Flow Control for the friends’ networks. In this paper we reinforce this model by analyzing its strength in terms of OSN features. We simulate attack scenarios carried out by a community of malicious users that attempt to fake the OSN features of the model. We analyze the attack of an alleged trustworthy clique of adversaries and show the futility of such an attack, due to the strength of the model’s parameters and combination of Trust, Access Control and Flow Control. We also demonstrate the robustness of the model when facing an optimized attack, which carefully selects the best network nodes to compromise, as determined by the minimal vertex cover algorithm.