Edith Law

Date: 22 February 2021

Time: 16:00 (CET)

Title: Crowdsourcing Medical Time Series Annotation: Expertise, Ambiguity and Human-AI Collaboration

Abstract: Many crowdsourcing contexts involve tasks that are short and simple, requiring little expertise or context. In this talk, I will discuss projects in which we tackle the problem of crowdsourcing medical time series annotation, addressing questions such as: To what extent can we engage non-experts to annotate medical data? What tools can we design to enable small groups of experts to collaboratively handle ambiguous edge cases? What are the various roles that AI can play in medical crowdsourcing systems? We will describe two platforms - MechanicalHeart and CrowdEEG - and findings from several studies that provide insights into how we can design systems that coordinate human and machine intelligence to tackle these difficult annotation problems.

Speaker Biography: Edith Law is an Associate Professor at the David R. Cheriton School of Computer Science at University of Waterloo. Her research focuses on how people can enhance intelligent systems (e.g., human-in-the-loop systems, crowdsourcing) as well as how people can make sense of intelligent systems, including issues related to transparency, engagement, trust and collaboration. She is interested in developing technologies that leverage the AI-people partnership to tackle more complex problems in business, science and medicine.

She is part of the Human Computer Interaction Lab. Her work is funded by NSERC Discovery Grant, NSERC-CIHR Collaborative Health Research Project (CHRP) as well as the CFI-JELF program.

Previously, she was a CRCS postdoctoral fellow at Harvard University. She graduated with a Ph.D. in Machine Learning from Carnegie Mellon University in 2012.

Homepage: http://edithlaw.ca/