Unpacking Public Dissatisfaction: Topic Modeling of Online News Comments on Spent Nuclear Fuel and High-Level Radioactive Waste Repositories in South Korea

As South Korea faces challenges in securing long-term solutions for spent nuclear fuel (SNF) and high-level radioactive waste (HLW) disposal, this study examines public discourse surrounding these issues through a large-scale analysis of online user comments. Unlike previous studies that relied on structured survey data, we apply an unsupervised natural language processing technique, latent dirichlet allocation topic modeling, to analyze approximately 21 000 highly liked comments posted under news articles related to SNF and HLW repositories from 2020.The results show that public distrust toward repository siting is deeply rooted in perceptions of regional inequity and political dissatisfaction, rather than technical concerns alone. In particular, comments are frequently associated with distributive injustice, where rural areas are seen as bearing the burden of energy infrastructure for the benefit of metropolitan regions. Furthermore, many other comments had little relevance to the issue of SNF and HLW and instead reflected politically and emotionally charged sentiments, often functioning as platforms for uncivil expression. By examining how the public spontaneously engages with radioactive waste issues in online spaces, this study provides new insights into the underlying causes of public opposition and highlights the importance of distributive fairness in shaping future radioactive waste policy.