Research talks: An investigation into user expectations for differential privacy
Differential privacy is widely regarded as a gold standard for privacy-preserving computation over users’ data. A key challenge with differential privacy is that its mathematical sophistication can be difficult to communicate to users, leaving them uncertain about whether they are protected. In this session, we will hear from two leading experts on how we can more effectively communicate differentially private guarantees to users, and how users might interpret these guarantees. Professor Rachel Cummings from Columbia University shares her work in exploring users’ privacy expectations related to differential privacy, shining a light on users’ concerns about differentially private computations and their willingness to contribute data to these systems. Following this presentation, Professor Jessica Hullman from Northwestern University shares her work on Visualizing Privacy (ViP), an interactive interface that visualizes relationships between privacy parameters, computation accuracy, and disclosure risk.
- Event:
- Research Summit 2022
- Track:
- Building User Trust Through Privacy, Identity, and Responsible AI
- Date:
-
-
-
-
Melissa Chase
Principal Researcher
-
-
