Ka I Chan she/her/hers

I am a first-year Ph.D. student at University of Michigan School of Information, advised by Professors Sun Young Park and Mark W. Newman. Before Michigan, I earned a master's degree in Data Science and Information Technology and a bachelor's degree in Psychology from Tsinghua University, where I was advised by Professors Yuanchun Shi and Yuntao Wang.

My research interests lie at the intersection of Human-Computer Interaction, Health Informatics, and Human-Centered AI. I study how AI-powered systems can support people's health and wellbeing in sensitive contexts, focusing on when and how these systems should intervene, and what makes them adaptive, supportive, and acceptable.

Ka I Chan
Recent News
Mar 2026"Capturing Chewing and Swallowing with Earables: A Multimodal Dataset Across Contexts" accepted at CHI '26 Workshop.
Nov 2025"Human and Algorithmic Visual Attention in Driving Tasks" accepted at npj Artificial Intelligence, from my internship at Tsinghua AIR.
Sep 2025"A Review of Behavioral Closed-Loop Paradigm from Sensing to Intervention for Ingestion Health" accepted at IMWUT '25.
Aug 2025Started my Ph.D. at UMSI with a first-semester fellowship!

Research

* denotes equal contribution
Full list on Google Scholar →
AI-Mediated Communication for Intimate Relationships Healthcare & Aging Sensing & Intervention for Eating Behavior

AI-Mediated Communication for Intimate Relationships

TOCHI (under review)
SpeakSoftly: Scaffolding Nonviolent Communication in Intimate Relationships through LLM-Powered Just-In-Time Interventions
Ka I Chan, Hongbo Lan, Jun Fang, Yuntao Wang, Yuanchun Shi
An LLM-powered just-in-time intervention system that scaffolds nonviolent communication during couple conflicts, focusing on when and how to intervene in sensitive interpersonal contexts.
PDF
In Preparation
Exploring AI Design Opportunities for Conflict Avoidance in Intimate Relationships
Jun Fang*, Ka I Chan*, Hongbo Lan, Yuntao Wang, Yuanchun Shi
An empirical study using interviews and co-design to explore how AI can support couples who tend to avoid conflicts, identifying design directions for adaptive communication support.

Healthcare & Aging

Ongoing
Designing LLM-based Chatbots for Older Adults' Post-Treatment Cancer Transition
Ka I Chan et al.
A storyboard-driven interview study investigating how LLM-based chatbots can support older adults treated for cancer during their post-treatment transition, examining design considerations across illness management, identity reconstruction, and daily life reintegration. 
International Journal of Human-Computer Studies
Beyond digital privacy: Uncovering deeper attitudes toward privacy in cameras among older adults
Weiwei Zhang, Jianing Yin, Ka I Chan, Tongxin Sun, Tongtong Jin, Jihong Jeung, Jiangtao Gong
A mixed-methods study revealing that older adults' privacy concerns with fall detection cameras extend beyond data protection to encompass dignity and perceived control over their aging identity.
PDF DOI
CHI 2025
The Odyssey Journey: Top-Tier Medical Resource Seeking for Specialized Disorder in China
Ka I Chan, Siying Hu, Yuntao Wang, Xuhai Xu, Zhicong Lu, Yuanchun Shi
A qualitative study using Actor-Network Theory to reveal how hemifacial spasm patients in China navigate social networks and technology to overcome information asymmetries and access top-tier medical resources.
PDF DOI
UIC 2024
Automated Grading Hemifacial Spasm Using Smartphone Cameras
Ka I Chan*, Bo Hei*, Linghao Meng*, Ruen Liu, Yuntao Wang, Chang Chen, Qingpei Hao, Yuanchun Shi
A smartphone-based AI system that detects and grades hemifacial spasm severity through facial keypoint analysis, achieving 88% detection accuracy aligned with clinical grading standards.
PDF DOI

Sensing & Intervention for Eating Behavior

IMWUT (under review)
Earinter: A Closed-Loop System for Eating Pace Regulation with Just-in-Time Intervention
Jun Fang, Ka I Chan, Xiyuxing Zhang, Yuntao Wang, Mingze Gao, Leyi Peng, Jiajin Li, Zihang Zhan, Zhixin Zhao, Yuanchun Shi
A system that uses commodity earbuds' bone-conduction sensing to detect chewing and swallowing, and deliver theory-grounded just-in-time interventions for healthier eating pace.
PDF
IMWUT 2025
A Review of Behavioral Closed-Loop Paradigm from Sensing to Intervention for Ingestion Health
Jun Fang*, Yanuo Zhou*, Ka I Chan, Jiajin Li, Zeyi Sun, Zhengnan Li, Zicong Fu, Hongjing Piao, Haodong Xu, Yuntao Wang, Yuanchun Shi
A systematic review of 136 studies proposing a behavioral closed-loop paradigm that integrates sensing, reasoning, and intervention for adaptive ingestion health systems.
PDF DOI
CHI 2026 Workshop
Capturing Chewing and Swallowing with Earables: A Multimodal Dataset Across Contexts
Jun Fang*, Ka I Chan*, Xiyuxing Zhang*, Yuntao Wang, Zihang Zhan, Zhixin Zhao, Yuanchun Shi
A multimodal dataset capturing chewing and swallowing events using commodity earables across noisy and quiet eating environments.
PDF