| Title |
A Comparative Analysis of the Government-General Building of Korea and the Gyeongseong Municipal Building Through Text Mining of Architectural Discourse |
| Authors |
이수연(Lee, Sooneon) ; 조은진(Jo, Eunjin) ; 이경아(Lee, Kyungah) |
| DOI |
https://doi.org/10.5659/JAIK.2026.42.10.243 |
| Keywords |
Government-General Building of Korea; Gyeongseong Municipal Building; Text Mining Analysis; Network Analysis; Cluster Analysis; Chosen to Kenchiku |
| Abstract |
Using sentence-level co-occurrence networks and Louvain community detection, this study compares the discourse surrounding two government
buildings completed in Gyeongseong in 1926. Analysis of two special issues of Chosen to Kenchiku identified eight clusters for the
Government-General Building of Korea and nine for the Gyeongseong Municipal Building. Although both buildings were discussed in terms
of construction, space, materials, and facilities, the discourse on the Government-General Building emphasized construction systems, material
production, infrastructure, and spatial hierarchy, presenting the building as a technical monument of colonial governance. In contrast, the
discourse on the Gyeongseong Municipal Building focused on decision-making actors, trade- and contractor-based construction, finishes,
individual devices, and floor- and function-based spaces, presenting it as a modern administrative facility. These differences demonstrate how
text mining can complement qualitative architectural history by revealing distinct patterns in architectural discourse. |