Have a UROP opening you would like to submit?
Please fill out the form.
A Mobile Game for Misinformation Detection
Kiran Garimella: firstname.lastname@example.org
On WhatsApp, the most popular messaging platform with over 2 billion active users, a large fraction of the content is shared through images. Understanding the spread of (mis)information on such platforms, in non textual modalities is important. Most work to date in addressing the problem of detecting misinformation primarily focuses on text. Detecting misinformation in images is a hard task, and is usually done by expert fact checkers or journalists, which does not scale well. In this project, we plan to build and deploy a game with a purpose to address this issue, thereby allowing us to utilize the wisdom of crowds to crowdsource fact checking images, and, in the process, collect a large scale high-quality labelled dataset that helps to investigate image-based misinformation. The primary task of the project would be to develop this game as an Android app or a mobile website. Key research questions include how to best design the game and aggregate the players' choices to utilize the wisdom of crowds. We will analyze the resulting data to assess different approaches.