Relationship Privacy Protection for Mobile Social Network

摘要

Recent scientific results have shown that location-based social networks (LBSNs) can be used to automatically and accurately predict even highly sensitive personal attributes (such as social relationships). To avoid the adversary inferring users' relationship privacy, a relationship privacy protection schema is proposed to protect the user privacy in mobile social network. An algorithm of location privacy protection is designed to calculate the probability of relationship privacy risk, which is used to evaluate whether the users' check-in meet the privacy requirement. Extensive experiments on a real dataset show that our method can effectively protect users' relationship privacy.