Digital Yŏm Sang-sŏp: Stylometry of His Fiction

Event time: 
Thursday, April 14, 2022 - 7:00pm to 8:30pm
Online () See map
Event description: 

Yŏm Sang-sŏp (1897-1963) is a major fiction writer and critic in South Korea. Famous for balancing between right- and left-wing ideologies and for a realistic writing style packed with witty punchlines, he portrayed Korean society during and after the Colonial Period (1910-1945). Unlike many Korean writers who relocated to the North, voluntarily or otherwise, or who died during and soon after the Korean War (1950-1953), Yŏm was an eyewitness to the formation and development of Korean literature in the South. It is thus unfortunate that we have yet to analyze his body of work as a whole. We have not even confirmed how many works he has left, nor have we examined the changes in his writing style before and after liberation. Here, we digitize as many of his fictional stories as possible and examine them via stylometry, a statistical analysis of various patterns in his literary style. Overall, our distant reading of word frequency and co-occurrence, text similarity and other formal features aims to compare trends in the narrative structure, such as characterization and plot, in Yŏm’s colonial and post-colonial works. As far as we are concerned, this is going to be the first attempt to approach Yŏm’s writings through computational stylistics in Korea’s literary studies.
Jae-Yon Lee is an associate professor of modern Korean literature at Ulsan National Institute of Science and Technology (UNIST) in South Korea. His research interests include twentieth-century Korean fiction and criticism, periodical studies, and digital literary studies. He has published collaborative works on network analysis, corpus analysis, and word embedding. He recently translated Franco Moretti’s Graphs, Maps, Trees into Korean and is working on a book manuscript that analyzes Korean periodicals through quantitative approaches.
Namgi Han is a postdoctoral fellow of informatics at Ulsan National Institute of Science and Technology (UNIST) in South Korea. In 2021, he graduated from the Graduate University for Advanced Studies, SOKENDAI, in Japan. His research interests include natural language processing and neural network interpretability. He has recently published papers examining the performance of neural networks by applying linguistic knowledge to question answering systems. He currently assists with teaching digital humanities and AI-generated writing courses at UNIST.

Jae-Yon Lee - Associate Professor of Modern Korean Literature, Ulsan National Institute of Science and Technology; Namgi Han - Postdoctoral Fellow of Informatics, Ulsan National Institute of Science and Technology