Privacy-preserving precise profile matching in mobile social network

Abstract

Profile matching is a key feature in mobile social networking applications, where friends are recommended to requesters by comparing the similarity of attributes between them.However, users’ personal information is exposed to the risk of privacy disclosure in the process of profile matching.The existing solutions exist some issues such as inaccurate matching results or inability to meet users’ requirements for multi-dimensional privacy protection.Baesed on this, a privacy-preserving precise profile matching (P3M) scheme was proposed, which allowed users to flexibly set the matching range of attributes and distances according to their requirements.The Paillier encryption was utilized to ensure data security of users, and a secure dot product protocol was designed to achieve secure ciphertext comparison of user attributes and query ranges.The P3M realized multi-dimensional privacy-preserving of users including user feature attributes and location attributes.Finally, the correctness and security of P3M scheme were analyzed and proved in detail, and extensive experimental results verified the effectiveness and efficiency of P3M scheme.

Keywords

• privacy-preserving;mobile social network;profile matching;order-preserving encryption