Privacy 2.0

Protection of privacy is a major concern for users of social web applications, including on-line social networks. Personal security breaches or loss of credibility and reputation can result when published information reaches unintended recipients. Although most online social networks now offer fine-grained controls of information sharing, these are rarely used, both because their use imposes additional burden on the user and because the control settings are too complex for an average user to handle. To mitigate this problem, we propose an intelligent privacy manager that partially automates the assignment of sharing permissions, taking into account the content of the information published and user’s high-level sharing policies. At the core of our contribution is a novel privacy policy language which explicitly accounts for social web concepts. The manager employs named entity recognition algorithms to annotate sensitive parts of published information and an answer set programming system for evaluating user’s privacy policies and determining the list of safe recipients. We have implemented a prototype of the manager on the Facebook platform. On a small test scenario, the system reached the F-measure value of 0.831 in correctly recommending safe recipients.

Contact: Michal Jakob

Facebook Intelligent Privacy Assistant from Agent Technology Center.





Michal Jakob (project leader), Zbynek Moler, Ondrej Pluskal, Roman Vaculin



Supported by Google Research Award.