DxD Lab

The DxD Lab is an interdisciplinary research group within the Industrial Design Department at KAIST. We explore design from, with, and by data to develop technologies that are more sensitive to human concerns and behaviors. Our research aims to build a more inclusive society by understanding the social impact of data and AI and by making technology accountable and accessible. Our current projects are developing human-centered and participatory approaches to designing algorithmic systems that are transparent, reliable, and responsive to community values and needs.

Research areas

Inclusive Algorithmic Experiences
Algorithms embedded in recent systems have inherent risks of providing non- or less-inclusive experiences to marginalized populations such as users with disabilities. We seek to design ways to foster an inclusive AX that could support the values of various users.

Human-Centered AI for Mental Health
Can AI empower individuals to be their own mental health agent? We examine AI’s potential for our mental health and propose ML/NLP applications that can support informed health decision. We also design novel methods to uncover underlying users’ perceptions and behaviors associated with algorithmic systems.

Data Humanism
Instead of focusing on the numbers, technologies, or algorithms around data, we should make room for a more personal approach that allows for context, serendipity, and even imperfections. We design approaches to synthesize data and storytelling in innovative ways to create visions of people, ideas, organizations, culture, and society.

news

Sep 1, 2022	MindScope has been featured in KAIST breakthrough newsletter.
Sep 1, 2022	Kwangyoung Lee joined DxD lab as a Ph.D. student 🎉
Sep 1, 2022	Sooyeon Ahn joined DxD lab as a Master’s student 🎉

Key publications

Prediction for Retrospection: Integrating Algorithmic Stress Prediction into Personal Informatics Systems for College Students’ Mental Health

Kim, Taewan, Kim, Haesoo, Lee, Ha Yeon, Goh, Hwarang, Abdigapporov, Shakhboz, Jeong, Mingon, Cho, Hyunsung, Han, Kyungsik, Noh, Youngtae, Lee, Sung-Ju, and Hong, Hwajung

CHI ’22, Apr 2022

Abs HTML

Reflecting on stress-related data is critical in addressing one’s mental health. Personal Informatics (PI) systems augmented by algorithms and sensors have become popular ways to help users collect and reflect on data about stress. While prediction algorithms in the PI systems are mainly for diagnostic purposes, few studies examine how the explainability of algorithmic prediction can support user-driven self-insight. To this end, we developed MindScope, an algorithm-assisted stress management system that determines user stress levels and explains how the stress level was computed based on the user’s everyday activities captured by a smartphone. In a 25-day field study conducted with 36 college students, the prediction and explanation supported self-reflection, a process to re-establish preconceptions about stress by identifying stress patterns and recalling past stress levels and patterns that led to coping planning. We discuss the implications of exploiting prediction algorithms that facilitate user-driven retrospection in PI systems.
“It’s not wrong, but I’m quite disappointed”: Toward an Inclusive Algorithmic Experience for Content Creators with Disabilities

Choi, Dasom, Lee, Uichin, and Hong, Hwajung

CHI ’22 Apr 2022

Abs HTML PDF

YouTube is a space where people with disabilities can reach a wider online audience to present what it is like to have disabilities. Thus, it is imperative to understand how content creators with disabilities strategically interact with algorithms to draw viewers around the world. However, considering that the algorithm carries the risk of making less inclusive decisions for users with disabilities, whether the current algorithmic experiences (AXs) on video platforms is inclusive for creators with disabilities is an open question. To address that, we conducted semi-structured interviews with eight YouTubers with disabilities. We found that they aimed to inform the public of diverse representations of disabilities, which led them to work with algorithms by strategically portraying disability identities. However, they were disappointed that the way the algorithms work did not sufficiently support their goals. Based on findings, we suggest implications for designing inclusive AXs that could embrace creators’ subtle needs.
Understanding University Students’ Experiences, Perceptions, and Attitudes Toward Peers Displaying Mental Health–Related Problems on Social Networking Sites: Online Survey and Interview Study

Kim, Taewan, and Hong, Hwajung

JMIR Mental Health ’21, Apr 2021

HTML PDF