I am a PhD Candidate at the Leiden Centre for Digital Humanities and the Leiden Institute for Area Studies at Leiden University. I study Korean history with digital methods, focusing on computer vision to read visual sources at scale and in detail. From June 2024 to March 2025, I was a Junior Research Fellow at the Kyujanggak Institute for Korean Studies at Seoul National University.
My work brings together quantitative image analysis and close historical reading. I use feature extraction, pattern matching, and visual clustering to trace how images circulated, how techniques and styles moved between workshops, and how visual conventions shaped readers’ experiences. The goal is to build tools and datasets that make visual materials first-class evidence in Korean historical research.
In practice, this means assembling digitized magazines and ephemera, curating metadata and annotations, and building reproducible pipelines that move from page images to interpretable results. I am especially interested in how computational approaches can surface patterns that are hard to see with close reading alone—while still remaining accountable to the archive, its gaps, and its biases.
Focus
I focus on visual methods for Colonial Korean materials that connect image-level evidence to historical questions. This includes training and evaluating models for layout, typography, and motif detection; constructing embeddings to compare imagery across publications; and aligning these quantitative traces with known production networks and readerships. A core part of my approach is building transparent, well-documented workflows that others can reuse or audit.
I also work on integrating computational results with archival description. Many historical collections have uneven metadata; I aim to enrich these records through semi-automated annotation and active learning, so that improved descriptions can circulate back to libraries, archives, and future researchers.
Current projects
Visual identities of Colonial Korean print shops: I am tracing techniques, type design, and recurring motifs across magazines to identify shop-specific signatures and changes over time. This involves linking visual features to issues, publishers, and known printers to reconstruct networks of production.
Consumer culture through advertisements: I am building a large-scale corpus of Colonial-era ads and using visual analysis to study themes, product categories, and persuasive strategies. By following visual campaigns across time and venues, the project explores how consumer imaginaries took shape and which audiences they targeted.
Background
Before my PhD, I modeled topics in Colonial Korean magazines to quantify leftist discourse in Kaebyŏk 개벽, which highlighted the value—and limits—of text-only approaches to visual periodicals. I then explored whether computer vision could attribute magazines to specific print shops, developing first-pass methods and benchmarks that now underpin my dissertation.
From June 2024 to March 2025, I was a Junior Research Fellow at the Kyujanggak Institute for Korean Studies (SNU). There, I deepened collaborations with archivists and scholars, expanded corpora, and refined pipelines for working at the intersection of archival description and visual analysis.
Roles
PhD Candidate ∙ Digital Humanities & Korean Studies ∙ Leiden University Researcher ∙ Leiden HumAN Junior Research Fellow (June 2024 – March 2025) ∙ Kyujanggak Institute for Korean Studies
Teaching
I teach the workshop series Deep Learning for the Humanities, which introduces fundamentals of neural networks, practical model training, dataset curation, and ethical considerations for humanities data. I also serve as T.A. for NLP for the Humanities, supporting assignments that bridge linguistic methods with real-world cultural data.
Education
M.A. Asian Studies (Korea Focus) ∙ Leiden University ∙ 2022 B.A. Korean Studies ∙ Leiden University ∙ 2021