- Computational Social Science
- Critical Data Studies
- Data Science
- Economics and Information
- Education Technology
- Ethics, Law and Policy
- Human-Computer Interaction
- Human-Robot Interaction
- Incentives and Computation
- Infrastructure Studies
- Interface Design and Ubiquitous Computing
- Natural Language Processing
- Network Science
- Social Computing and Computer-supported Cooperative Work
- Technology and Equity
Yea-Seul Kim is a Ph.D. candidate in the Information School at the University of Washington, advised by Dr. Jessica Hullman. Her research interests lie in the intersection between Visualization, HCI and Data Science. Her work aims to develop tools and algorithms that can help make data more comprehensible to people of varying levels of expertise. She is especially interested in how more formal models of belief updating can be used to transform how we design and evaluate visualization systems and techniques. She received a master’s degree in Human-Centered Design and Engineering at the University of Washington and a bachelor’s degree in Applied Statistics at Yonsei University in Korea. Her work received a best paper award at the CHI Conference on Human Factors in Computing Systems in 2017, and she was named a Rising Star in EECS in 2019.
Talk: "Designing Beliefs-driven Interactions with Data"
This talk will be held virtually via Zoom.
Abstract: As society becomes increasingly data-driven, people encounter numerical estimates about topics ranging from socially relevant phenomenon (e.g., job growth statistics) to personal informatics (e.g., heart rate data) on a daily basis. Visualizations effectively summarize data to help people more easily grasp the phenomenon that data represents. Evidence from cognitive psychology, behavioral economics, and related fields demonstrate that beliefs play a key role in how people interpret and comprehend new information. However, research in Visualization and HCI lacks design patterns and techniques for integrating users' beliefs into human-data interactions, as well as formal models for evaluating the interactions.
In this talk, I will show how visualization research can fill this gap by combining novel interfaces for interacting with one’s beliefs with models of Bayesian inference to quantify and evaluate people’s data interpretation processes in light of their beliefs. I will show how accounting for beliefs in visualization interpretation and interaction provides deeper insight into how we should design and evaluate visualizations. I will introduce several interactive graphical techniques I developed to elicit people’s beliefs by emulating how people reason about uncertainty. I will also describe interactive applications that I have been developing that use the user's prior beliefs to personalize presentations of data to improve uncertainty comprehension, or that generate representations based on cognitively familiar metaphors to make data relatable.