Title:Camera Fingerprint: A New Perspective for Identifying User's Identity

Abstract:Identifying user's identity is a key problem in many data mining applications, such as product recommendation, customized content delivery and criminal identification. Given a set of accounts from the same or different social network platforms, user identification attempts to identify all accounts belonging to the same person. A commonly used solution is to build the relationship among different accounts by exploring their collective patterns, e.g., user profile, writing style, similar comments. However, this kind of method doesn't work well in many practical scenarios, since the information posted explicitly by users may be false due to various reasons. In this paper, we re-inspect the user identification problem from a novel perspective, i.e., identifying user's identity by matching his/her cameras. The underlying assumption is that multiple accounts belonging to the same person contain the same or similar camera fingerprint information. The proposed framework, called User Camera Identification (UCI), is based on camera fingerprints, which takes fully into account the problems of multiple cameras and reposting behaviors.