Byunghwee Lee (이병휘) is a computational social scientist and data scientist who develops interpretable, quantitative frameworks to uncover how human beliefs, behavior, and creativity are structured and how they evolve.

I am currently a Postdoctoral Research Associate at the University of Virginia School of Data Science, working with Prof. Yong-Yeol Ahn. My research integrates machine learning, large language models, and network science to study human beliefs, cultural expression, and collective social dynamics using large-scale behavioral and cultural data.

• Email: wzn3hf@virginia.edu
• CV: cv.pdf
• Office: School of Data Science, University of Virginia, Charlottesville, VA.
• Research Area: Computational social science, Cultural analytics, Belief system, Representation learning, LLMs and multimodal AI

Research Focus#

• Interpretable belief representation and belief dynamics with large language models
• Computational analysis of art history, creativity, and multimodal cultural data
• Network science approaches to collective behaviors in online media and social systems

Explore#

Welcome to my homepage! You can find my research topics, publications, news, and essays below.

• 🧐 Browse current research
• 🗺️ Visit visualization
• 📜 See publications and talks
• 🌎 Explore my Hwiki (thoughts, tips, essays, etc.)

Previous Positions#

• Postdoctoral Researcher, Luddy School of Informatics, Computing and Engineering, Indiana University Bloomington, 2023-2025
  • Advisors: Jisun An, Haewoon Kwak, and Yong-Yeol Ahn
• Senior Researcher, Institute of Basic Science, Sungkyunkwan University, 2022-2023
  • Advisor: Beom Jun Kim
• Researcher / Curator, National Science Museum, 2021-2022
• Postdoctoral Researcher, Natural Science Research Institute, KAIST , 2021
  • Advisor: Hawoong Jeong

Education#

• PhD in Physics, KAIST, 2021
• BS in Physics, KAIST, 2013

News#

• 2026-04-07: New paper released: When the Classroom Disappeared: The Paradox of Assortativity in Co-Enrollment Networks on Social Constellations: A World Perspective
• 2026-03-21: Added a Visualization page to my website—more to come soon!
• 2026-03-16 : New preprint released: LLMs Can Infer Political Alignment from Online Conversations on arXiv
• 2025-11-18 : Received the Datapalooza Best in Show Award at the University of Virginia School of Data Science News
• 2025-07-31 : Published Network analysis reveals news press landscape and asymmetric user polarization in Physica A
• 2025-06-04 : Published A semantic embedding space based on large language models for modelling human beliefs in Nature Human Behaviour