CS Colloquium

Time and Location:

Speaker:

Link:

Technical approaches for protecting data privacy have immense potential to safeguard sensitive data and thus create critical infrastructure for responsible data science and AI. However, these emerging approaches are largely disconnected from on-the-ground practice, limiting their adoption and opportunities for real-world feedback. Furthermore, protecting privacy at scale requires people with distinct roles, responsibilities, and needs to all integrate novel practices into their already-challenging routines. In this talk, I will describe my approach to enabling people across the data ecosystem to protect data privacy & security by presenting three projects that develop interactive, visual tools that align differential privacy—a state-of-the-art approach to privacy-preserving data science and AI—with how data curators and data subjects naturally reason about data. This includes research that formed the seeds for a $1.5m NSF grant I co-authored and informed a tool that the U.S. National Institute of Standards and Technology has formally proposed hosting. Finally, I will discuss promising directions for future work that continue to leverage a human-centered strategy to advance data privacy & security.