Hi, I am Soohwan 👋
About Me
I am a Ph.D. candidate in the Expressive Computing Lab at Department of Design, UNIST, specializing in Human–Computer Interaction (HCI) and Human-centered AI.
My research investigates how AI systems reshape group dynamics such as group decision-making, social influence, and power dynamics, drawing on theories from social psychology and designing interactive AI systems grounded in empirical evidence. Rather than focusing on isolated interactions between a single user and an AI system, my work advances a group-level perspective, examining how people experience, interpret, and negotiate influence when AI agents participate in collective settings.
Research Focus
My research centers on AI-mediated group dynamics, with particular attention to how social influence unfolds over time and how power asymmetries shape collective outcomes. I study these phenomena through a combination of experimental studies, system design, and theory-driven analysis.
Key themes in my work include:
- AI Social Influence in Groups
Investigating how AI agents exert normative and informative influence, and how people respond through compliance, conversion, or resistance in multi-agent settings. - Group-centered AI & Group Experience (GX)
Proposing conceptual frameworks that extend Human-centered AI beyond the individual level to account for collective experience, coordination, and conflict. - Minority Support and Dissenting AI Interventions
Designing LLM-powered systems that support minority perspectives through counterargument, reframing, and mediation, particularly in power-imbalanced group decision-making. - Reflective Interfaces for Group Sensemaking
Exploring visualization and interaction techniques that help groups reflect on value conflicts, participation asymmetries, and influence trajectories during collaboration.
Current Projects
- LLM-powered Minority Support Systems
AI-mediated dissent and mediation strategies that amplify minority viewpoints and promote critical discussion in group decision-making. - Multi-agent Social Influence Dynamics
Experimental investigations of how majority, minority, and diffusion patterns of AI influence shape opinions and confidence over time. - Group Reflective Dashboard
A visualization-based system that surfaces latent value conflicts, influence trajectories, and participation asymmetries to support post-hoc and in-situ group reflection.
Research Vision
Across these projects, my broader goal is to design Group-centered AI systems that do not simply optimize group outcomes, but help people understand how those outcomes emerge, whose values are reflected, and how influence is exercised within AI-augmented collectives.
Personal Interests
In my free time, I enjoy running and exploring a diverse range of films and literature. I particularly love movies directed by Quentin Tarantino, Christopher Nolan, and Denis Villeneuve. However, some of my favorite films include ‘Sicario’, ‘The Matrix’, and ‘Everything Everywhere All at Once’. I also have a particular interest in reading books and articles on cross-cultural psychology, such as ‘Thought of Geography’.