Towards understanding privacy risks in online social networks

Student thesis: Doctoral thesis

Abstract

Online Social Networks (OSNs) are now one of the most popular services on the Internet. When these lines were written, there were four OSN sites in the Alexa's top ten global ranking and the most used OSNs were having hundreds of millions of daily active users. People use OSNs to share all kinds of contents: from personal attributes (like names, age, or gender), to location data, photos, or comments. Moreover, OSNs are characterized by allowing its users to explictly form relationships (e.g. friendship). Additionally, OSNs include not only information the users conscientiously post about themselves, but also information that is generated from the interaction of users in the platform. Both the number of users and the volume of data shared make privacy in OSNs critical. This thesis is focused on studying privacy related to OSNs in two different contexts: crawling and learning. First, we study the relation between OSN crawling and privacy, a topic that so far received limited attention. We find this scenario interesting because it is affordable for even a low-budget attacker. Second, we study how to extract information from the relationships OSN users form. We then expand our findings to other graph-modeled problems.
Date of Award17 May 2016
Original languageEnglish
Supervisor Bart Preneel (Director), Maria Claudia Díaz Martínez (Director) & Jordi Herrera Joancomarti (Director)

Keywords

  • Online social networks
  • Privacy
  • Realational learning

Cite this

'