Chenhao Tan, University of Colorado Boulder
Host: Dragomir Radev
Coffee available at 3:45
Title: Towards Human-Centered AI: Understanding Language in Social Contexts
Language plays a central role in shaping human behavior and opinions. However, understanding language in social contexts is challenging. For instance, it is well recognized that humans struggle with detecting deception. Such challenging tasks suggest that AI algorithms can discover new knowledge from data beyond simply emulating human behavior. This discovering mode gives rise to opportunities and challenges of augmenting humans with AI. In this talk, I will first present methods for advancing AI in the discovering mode: 1) deriving insights from a text corpus by developing a neural framework for incorporating document metadata such as tone; 2) identifying the effect of wording in information diffusion by controlling for confounding factors through natural experiments. I will then share our recent effort in developing best practices for incorporating AI into human decision making in such challenging tasks, using deceptive review detection as a testbed. Our results demonstrate a tradeoff between human performance and human agency and show that explanations of machine predictions can moderate this tradeoff.
Chenhao Tan is an assistant professor of computer science at the University of Colorado Boulder. He obtained his PhD degree in the Department of Computer Science at Cornell University and bachelor’s degrees in computer science and in economics from Tsinghua University. Prior to joining CU Boulder, he spent a year at the University of Washington as a postdoc. His research interests include human-centered AI, natural language processing, and computational social science. His work has been covered by many news media outlets, such as the New York Times and the Washington Post. He also won an NSF CRII award, a Salesforce research award, a Facebook fellowship, and a Yahoo! Key Scientific Challenges award.