AI-driven User Profiling
February 17, 2021
Xinru Page

AI-driven user profiling based on user data is a key component of providing a personalized experience on Facebook. However, public concern has grown around social media platforms profiling users to infer fine-grained interests and characteristics such as race, religion, or socioeconomic status, which can be used for secondary purposes such as targeted advertising. In fact, there have been recent public outcries over how such micro profiling has been used to sway political views, as well as concern over how people’s emotional and psychological well-being can be negatively affected. This project investigates how to leverage one of the most popular medium for engaging viewers, Youtube videos, to increase digital literacy and help users limit such profiling. This is an interdisciplinary effort drawing from Media and Film studies, Applied Artificial Intelligence, and Human-Computer Interaction.

Xinru Page recently received funding to increase digital literacy around user profiling.

Increasing Transparency of Fine-grained User Profiling: Deploying Visual Interventions for Users of Varying Digital Literacy
Sponsor: FCLTech (Facebook)

This project investigates how to educate people about how they are profiled on social media, the real world consequences, and how to take control of their privacy. This is an interdisciplinary collaboration drawing on theories from media and film studies, education, and applied artificial intelligence.