On analyzing and evaluating privacy measures for social networks under active attack

TL;DR

This paper investigates privacy violation measures in social networks under active attacks, providing theoretical insights and empirical analysis of real and synthetic networks to inform better privacy-preserving network design.

Contribution

It offers a theoretical condition for network design to prevent privacy violations and empirically analyzes privacy vulnerabilities in various real and synthetic social networks.

Findings

Networks without cycles require careful design to prevent privacy violations.

Empirical analysis reveals privacy vulnerabilities in real social networks.

Synthetic network models exhibit different privacy violation properties.

Abstract

Widespread usage of complex interconnected social networks such as Facebook, Twitter and LinkedIn in modern internet era has also unfortunately opened the door for privacy violation of users of such networks by malicious entities. In this article we investigate, both theoretically and empirically, privacy violation measures of large networks under active attacks that was recently introduced in (Information Sciences, 328, 403-417, 2016). Our theoretical result indicates that the network manager responsible for prevention of privacy violation must be very careful in designing the network if its topology does not contain a cycle. Our empirical results shed light on privacy violation properties of eight real social networks as well as a large number of synthetic networks generated by both the classical Erdos-Renyi model and the scale-free random networks generated by the Barabasi-Albert preferential-attachment model.