ShutterRoller: Preserving Social Network Privacy towards High-Speed Domain Gateway

Online Social Network (OSN) has become an important platform where people share their life and opinions nowadays, which is followed by security problems especially privacy leakage. Contents generated by OSN users can be browsed and shared by friends or even strangers, which may contain their privacy information and leads to a high risk of privacy leakage. However, traditional DPI mechanism, such as NIDS and NIPS, can hardly adapt to privacy detection and preservation in OSN because of their granularity limitation or poor performance. Privacy settings of OSN sites are seldom changed by users. In this paper, we proposed a social network privacy preserving system, ShutterRoller, based on domain gateway. By detecting the OSN traffic through gateway, ShutterRoller examines the existing of social features that may lead to privacy leakage, such as user behaviors and user generated contents. The work of ShutterRoller is accomplished with the following contributions: (1) a general network underlying framework for application layer data restoration with high throughput and low latency towards the high-speed network environment of gateway. (2) a customized OSN application layer inspection scheme to achieve fine-grained privacy leakage detection with a high accuracy. Experiments have been done by real network traffic, and indicate the high accuracy in fine-grained privacy leakage detection and high throughput as well as low latency in high-speed network.