LinkMirage: How to Anonymize Links in Dynamic Social Systems

摘要: Social network based trust relationships present a critical foundation for designing trustworthy systems, such as Sybil defenses, secure routing, and anonymous/censorshipresilient communications. A key issue in the design of is revelation users' trusted social contacts to an adversary-information that considered sensitive today's society. In this work, we focus on challenge preserving privacy contacts, while still enabling applications. First, propose LinkMirage, community detection algorithm anonymizing links topologies; LinkMirage preserves structures topology within communities. considers evolution topologies, minimizes leakage due temporal dynamics system. Second, define metrics quantifying utility time series topologies with anonymized links. We analyze provided by both theoretically, well using real world topologies: Facebook dataset 870K large-scale Google+ 940M find our approach significantly outperforms existing state-of-art. Finally, demonstrate applicability real-world applications reputation anonymity systems vertex anonymity. also prototype application can bootstrap privacy-preserving without cooperation OSN operators.