Leveraging Personalized Internet Services to Combat Online Trolling
This research is supported by the National Science Foundation.
Motivation and ApproachPublic opinion is of paramount importance in any society. It is thus not a surprise that many governments, political parties, and various other groups deploy tactics to influence public opinion on the Internet, a practice commonly referred to as trolling. Organized trolling has already become a serious problem, and some argue that it can have a profound impact on the society. While resolving versions of this problem in the past has been done in the context of individual systems, e.g., a social network, resolving the problem comprehensively at the Internet-scale is of paramount importance for the future of the Internet.
In this project, we ask if it is possible to utilize the ubiquitous online tracking of users, performed by numerous trackers associated with web site recommendations, ad networks, search engines, online social networks, mobile applications, etc., for the direct benefit of the users themselves as well as the numerous distributed systems that rely upon open membership. On one hand, users face a contradiction: despite the fact that almost every browser click made over the last decade has been monitored by numerous online trackers, the users often have a hard time proving their identity and uniqueness while using the Internet. On the other hand, many systems that rely upon open membership are often targets of online trolling. Traditional defenses against such attacks rely on trusted identities or strong assumptions about the structure of the social network or the behavioral patterns of attackers. However, requiring users to present trusted identities runs against the open membership that underlies the success of these online services in the first place, and the assumptions have been shown not to hold true for adaptive attackers.
We propose Co-opt, a system that enables the users to take direct advantage of the work online trackers do to record and interpret their behavior. Our key idea is to use the readily available personalized content, generated by online trackers in real-time, as a means to generate a unique user identity in a seamless and privacy-preserving manner. We propose to utilize such tracker-generated personalized content, submitted directly by the user, to construct a multi-tracker user-vector representation that is unique to the user, and then use this vector as a mechanism to counter online trolling. The main research objectives of this proposal, outlined in detail below, are to explore the fundamental properties of such user-vector representations, i.e., their construction, uniqueness, persistency, resilience, and utility in counter-trolling scenarios. The goal of this project is to design, implement, and evaluate such a counter-trolling service.
- Ph.D. Students:
- Undergraduate Students:
- Utilizing Web Trackers for Sybil Defense
M. Flores, A. Khan, M. Warrior, A. Mislove, and A. Kuzmanovic
Potential for Discrimination in Online Targeted Advertising
Till Speicher, Muhammad Ali, Giridhari Venkatadri, Filipe Nunes Ribeiro, George Arvanitakis, Fabricio Benevenuto, Krishna P. Gummadi, Patrick Loiseau, and Alan Mislove
In Proceedings of the Conference on Fairness, Accountability, and Transparency (FAT*'18), New York, NY, USA, Feb 2018.
Investigating Ad Transparency Mechanisms in Social Media: A Case Study of Facebook's Explanations
Athanasios Andreou, Giridhari Venkatadri, Oana Goga, Krishna P. Gummadi, Patrick Loiseau, and Alan Mislove
In Proceedings of the Network and Distributed System Security Symposium (NDSS'18), San Diego, CA, USA, Feb 2018.
Privacy Risks with Facebook's PII-based Targeting: Auditing a Data Broker's Advertising
Giridhari Venkatadri, Yabing Liu, Athanasios Andreou, Oana Goga, Patrick Loiseau, Alan Mislove, and Krishna P. Gummadi
In Proceedings of the IEEE Symposium on Security and Privacy (IEEE S&P'18), San Francisco, CA, USA, May 2018.