Sarah M. Brown
University of Rhode Island
Designing a Tool to Measure Perceptions of AI Fairness
Understanding the impact of AI on society requires understanding how people feel about their impact and what people, outside those building them, think AI should do. However, while performance of algorithms is measured in terms of a now field-standard set of performance metrics, these metrics and their interpretation are not standard knowledge for most people who are impacted by these systems. To express their thoughts reliably, non expert study participants need to understand what they are being asked. We hypothesize that with careful orientation, interactive data visualizations can decrease the barriers to understanding complex data that are necessary to ask opinions about it. We propose using interactive data visualizations as a data collection tool for understanding people's preferences. In this talk I will discuss the design and validation of the tool, including results from a preliminary study.
Sarah M. Brown is an Assistant Professor of Computer Science and director of the ML4STS Lab at the University of Rhode Island. Previously she was a Data Science Initiative Postdoc at Brown University affiliated to the Division of Applied Mathematics and hosted by Professor Bjorn Sandstede and she completed a Chancellor’s Postdoctoral Fellow in Computer Science at the University of California, Berkeley with faculty mentor Professor Mike Jordan. She completed a BS in Electrical Engineering with a minor in Biomedical Engineering in 2011 a MS in Electrical and Computer Engineering and a PhD in Electrical Engineering in 2016 advised by Jennifer Dy both at Northeastern University. Her graduate studies were supported by a Draper Laboratory Fellowship and a National Science Foundation Graduate Research Fellowship. Brown's other professional activities include teaching computational data analysis skills to researchers with The Carpentries and serving as the treasurer emeritus on the Women in Machine Learning, Inc Senior Advisory Council.