Social Computing Lab, advised by Wonjae Lee
Hyeonseung Kim
M.S. Student · Graduate School of Metaverse, KAIST
NLP · Explainable AI · Social Media Analysis
About
I am an M.S. student at Social Computing Lab, Korea Advanced Institute of Science and Technology (KAIST), advised by sociologist Wonjae Lee. My vision is a future where humans and AI coexist in harmony, and where AI is used to ease social conflict rather than intensify it. My research focuses on trustworthy and transparent AI and on understanding online information ecosystems, with the goal of fostering healthier public discourse and a less divided society.
Research Interests
- Explainability: I develop interpretable and human-centered AI methods that make complex machine learning systems more transparent, reliable, and accountable. My research focuses on understanding how AI models represent and reason about social and linguistic signals, enabling the detection and explanation of phenomena such as coordinated behavior, moral and emotional expression, and information dynamics in online environments.
- Social Impact: I study how online information environments shape human behavior and society. By combining computational social science, natural language processing, and large-scale behavioral analysis, I investigate how online discourse, moral emotions, and information operations influence public opinion, political participation, user engagement, collective behavior, and polarization.
News
2026.08
Paper accepted at The 2026 Symposium on Electronic Crime Research (eCrime 2026).
2026.08
Presented a poster at USENIX Security Symposium (USENIX Security '26) in Baltimore, MD.
2026.06
Paper accepted at USENIX Security Symposium 2026.
2026.04
Paper accepted at Frontiers in Oncology.
2026.01
Paper accepted at The Web Conference (WWW '26).
Education
Korea Advanced Institute of Science and Technology (KAIST)
Handong Global University
Data Analysis Lab
Experience
SKIA
- Developed medical augmented reality applications for surgical planning and visualization
- Built deep learning segmentation models for pelvic and hepatic vasculature detection from CT/MRI scans
- Worked across Computer Vision, iOS (Swift), and Unity development pipelines
Codestates
- Designed and delivered curriculum covering AI and machine learning fundamentals
Publications
Hyeonseung Kim, Jaehong Kim, Meeyoung Cha, Wonjae Lee. "Suspected Foreign Influence Activity in Korean News Discussions on YouTube." Symposium on Electronic Crime Research (eCrime 2026), 2026. To appear
Jaehong Kim*, Hyeonseung Kim*, Jiseon Kim, Alice Oh, Thorsten Holz, Wonjae Lee, Meeyoung Cha. "Cross-National Information Attacks: A Two-Decade Analysis of Troll Behavior in Korea." USENIX Security Symposium (USENIX Security '26), 2026. To appear
* Equal contribution
Hyeonseung Kim, Min Jin Jeong, Kyo Yeong Koo, Youn Jin Choi, Eun Seo Heo, Woohyun Nam, Sangyun Kang, U-Young Lee, Yi-Suk Kim, Keun Ho Lee, Chan-Ung Park. "Deep Learning Accurately and Reliably Segments Pelvic Vascular Structure in CT Scans of Gynecologic Cancer Patients." Frontiers in Oncology, 2026. https://doi.org/10.3389/fonc.2026.1687859
Seongchan Park, Jaehong Kim, Hyeonseung Kim, Heejin Bin, Sue Moon, Wonjae Lee. "Moral Outrage Shapes Commitments Beyond Attention: Multimodal Moral Emotions on YouTube in Korea and the US." The Web Conference (WWW '26), 2026. https://doi.org/10.1145/3774904.3792728