Research

How Audit Methodologies Impact Our Understanding of YouTube’s Recommendation Systems | ICWSM' 24

Sarmad Chandio, Daniyal Dar, Rishab Nithyanand $~$

Computational audits of social media websites have generated data that forms the basis of our understanding of the problematic behaviors of algorithmic recommendation systems. Focusing on YouTube, this paper demonstrates that conducting audits to make specific inferences about the underlying content recommendation system is more methodologically challenging that one might expect.

The Effect of Monetary Incentives on the Behavior of Online Content Moderators | Working

Online content moderators keep the internet clean, sometimes at the expense of their mental health. Focusing on Reddit moderators, we are trying to understand the effect of monetary incentives on the intrinsic motivation and work quality of these volunteer moderators.

Systematically Detecting Data Voids | Working

A user is most vulnerable when finding information online. Malicious actors fill up the voids on the internet leading vulnerable users to their misinformation bubbles. In this work we are designing infrastrucutres to systematically detect the exploitative voids.