Shujaat Mirza, Christina Pöpper (New York University)

Online social networks accumulate unprecedented amounts of data that continue to exist on user profiles long after the time of posting. Given that these platforms primarily provide a venue for people to connect and foster online friendships, the influence and the risks associated with longitudinal data may impact users and their reasons for using these platforms. To better understand these issues, we conducted two user studies of Facebook users analyzing the history of past postings w. r. t. to their perceived relevance, longitudinal exposure, and impact on the users’ befriending behavior. The studies give us a cross-cultural undergraduate student sample (n=89, campus study) and a Mechanical Turk sample of two cultural backgrounds from the US and India (n=209, MTurk study). Our findings reveal that a sizable group of participants consider their past postings irrelevant and, at times, embarrassing. However, participants’ awareness and usage of longitudinal privacy control features (e. g., Limit Past Posts) are limited, resulting in overexposure of their past postings and personal information. Importantly, we find support that these overexposed, yet irrelevant, past postings (of both participants and friend requesters) have the potential to influence users’ fundamental behavior on the platform: friend network expansion. Participants greatly valued friend requester’s past postings, particularly in the absence of prior personal interactions, but are influenced by their backgrounds (American users rely significantly more than their Indian counterparts on the requesters’ past postings for their befriending behavior). We close by discussing the implications of our findings on the future of longitudinal privacy controls.

View More Papers

Mondrian: Comprehensive Inter-domain Network Zoning Architecture

Jonghoon Kwon (ETH Zürich), Claude Hähni (ETH Zürich), Patrick Bamert (Zürcher Kantonalbank), Adrian Perrig (ETH Zürich)

Read More

POSEIDON: Privacy-Preserving Federated Neural Network Learning

Sinem Sav (EPFL), Apostolos Pyrgelis (EPFL), Juan Ramón Troncoso-Pastoriza (EPFL), David Froelicher (EPFL), Jean-Philippe Bossuat (EPFL), Joao Sa Sousa (EPFL), Jean-Pierre Hubaux (EPFL)

Read More

Oblivious DNS over HTTPS (ODoH): A Practical Privacy Enhancement...

Sudheesh Singanamalla*†, Suphanat Chunhapanya*, Jonathan Hoyland*, Marek Vavruša*, Tanya Verma*, Peter Wu*, Marwan Fayed*, Kurtis Heimerl†, Nick Sullivan*, Christopher Wood* (*Cloudflare Inc. †University of Washington)

Read More

Demo #2: Sequential Attacks on Kalman Filter-Based Forward Collision...

Yuzhe Ma, Jon Sharp, Ruizhe Wang, Earlence Fernandes, and Jerry Zhu (University of Wisconsin–Madison)

Read More