About Me
I am a postdoctoral researcher and Jang Young Sil Fellow (Excellence Track) at KAIST, working with Uichin Lee in the Interactive Computing Lab.
My goal is to advance human–computer interaction and build human-centered AI that understands people and that people can understand. I combine multimodal learning, models of human behavior, and human-centered evaluation to study AI across mental health, everyday behavior, and online social interaction.
My current interests follow two connected directions. (1) Longitudinal personalization: I study how personal AI can learn from everyday records, maintain and update memory, and adapt as people change.
(2) User simulation for agent evaluation and learning: I am interested in user digital twins and behavior-grounded simulation for evaluating and training personal agents. This includes reinforcement-learning environments informed by longitudinal user data, as well as representations that users can inspect and revise.
I received my Ph.D. in Artificial Intelligence from Hanyang University, advised by Kyungsik Han. From August 2019 to February 2020, I was a visiting scholar in the PIKE Lab at Pennsylvania State University, working with Dongwon Lee.
Recent News
Our paper on longitudinal mental health monitoring was accepted to NeurIPS 2026. I will present it in Sydney in December.
Sonju received the People’s Choice Student Challenge Award at UbiComp/ISWC 2026. I co-advised the project with Uichin Lee.
Selected for the NRF Basic Research Program B (Inclusive Track) as Principal Investigator.
Joined KAIST as a Jang Young Sil Fellow (Excellence Track).
Publications
* Equal Contribution † Corresponding author
Conference & Journal Papers
-
Dynamic Context Modeling for Longitudinal Mental Health Monitoring under Distribution Shift
NeurIPS 2026 To appear
-
“Can LLMs Persuade Humans with Deception?”: From a Deceptive Strategy Taxonomy to a Large-Scale Empirical Study
CHI 2026
-
TRIPLE: Theory-Driven Integration of Planned and Habitual Behaviors for LLM-based Personalization
AAAI 2026
-
Externalizing Social-Cognitive Structures for User Modeling: Toward Theory-Driven Profiling with LLMs
CIKM 2025
-
“I Don’t Know Why I Should Use This App”: Holistic Analysis on User Engagement Challenges in Mobile Mental Health
CHI 2025
-
PADO: Personality-induced multi-Agents for Detecting OCEAN in human-generated texts
COLING 2025
-
ICEv2: Interpretability, Comprehensiveness, and Explainability in Vision Transformer
International Journal of Computer Vision 2025
Online publication: November 2024 -
PREDICT: Multi-Agent-based Debate Simulation for Generalized Hate Speech Detection
EMNLP 2024
-
Integration of Global and Local Representations for Fine-grained Cross-modal Alignment
ECCV 2024
-
Label-aware Hard Negative Sampling Strategies with Momentum Contrastive Learning for Implicit Hate Speech Detection
ACL Findings 2024
-
An Empirical Study on Social Anxiety in a Virtual Environment through Mediating Variables and Multiple Sensor Data
CSCW 2024
-
Adversarial Normalization: I Can visualize Everything (ICE)
CVPR 2023
-
AI-Augmented Art Psychotherapy through a Hierarchical Co-Attention Mechanism
CIKM 2022
-
FashionQ: An AI-Driven Creativity Support Tool for Facilitating Ideation in Fashion Design
CHI 2021
-
FANCY: Human-centered, Deep Learning-based Framework for Fashion Style Analysis
WWW 2021
Workshop, Demo & Other Conference Papers
-
Sonju: A Context-Aware Teachable Agent for Eliciting Older Adults’ Situated Knowledge
UbiComp/ISWC 2026 (Student Challenge)
People’s Choice Student Challenge Award -
Designing Clinician Centered Explanations for Sensing-based Longitudinal Mental Health Monitoring
UbiComp/ISWC 2026 (XAI for U Workshop)
-
A Contactless and Non-Intrusive System for Driver’s Stress Detection
UbiComp/ISWC 2023 (Poster)
-
FashionQ: An Interactive Tool for Analyzing Fashion Style Trend with Quantitative Criteria
CHI 2020 (Late-Breaking Work)
Books & Reports
General-Purpose Artificial Intelligence (GPAI) Risk Management Framework
Telecommunications Technology Association (TTA), 2025.
Research report (in Korean). Contributing researcher.
Research Grants
NRF Basic Research Program B (Inclusive Track)
Mitigating Distribution Shift in Mental Health Prediction via Non-Parametric Multi-Agent Dynamic Context Modeling
Principal Investigator · National Research Foundation of Korea (NRF)
Sep 2026 – Aug 2029 · KRW 180 million over three years
Jang Young Sil Fellow (Excellence Track)
Development of a Multi-Agent-based Longitudinal Personalization Framework for User Mental Health Monitoring
Principal Investigator · Ministry of Science and ICT (MSIT) and KAIST
Apr 2026 – Dec 2026 · KRW 80 million annual research budget
NRF Ph.D. Fellowship
Research on Supporting Art Psychotherapy through Augmented Artificial Intelligence
Principal Investigator · National Research Foundation of Korea (NRF)
Jun 2021 – May 2023 · KRW 40 million over two years
Awards
- People’s Choice Student Challenge Award, UbiComp/ISWC (2026)
- Best Paper Award, Korea Data Mining Society (2025)
- Best Paper Award, Korea Data Mining Society (2024)
- Best Global Research Team Award, IITP (2021)
Academic Services
Reviewer · Artificial Intelligence (AI)
- WWW (2027)
- NeurIPS (2026)
- CVPR (2026)
- ARR (January, March, May 2026)
- DSAA (2025)
Reviewer · Human-Computer Interaction (HCI)
- CHI (2024, 2027)
- UIST (2026)
- IMWUT (2026)
- CHI LBW (2024–2025)
- CSCW (2024)
Student Volunteer
- CHI (2020)
Teaching
- Teaching Fellow, Human-Computer Interaction (Fall 2025)
- Co-lecturer, Creative Software Design (Fall 2021, 2022, 2023)
- Co-lecturer, Introduction to Software Design (Spring 2022, 2023)