Suicide Phenomena in South Korea from 2011 to 2020: Text Mining and Network Analysis of News Using Big Data

This study aimed to identify latent clusters underlying suicide phenomena in South Korea from 2011 to 2020, a period marked by the country's highest suicide rates. To achieve this, 12,570 news articles were collected from BIG KINDS, a news article database of the Korea Press Foundation, and analyzed using big data techniques. Text mining was applied to article titles using Textom, followed by CONCOR analysis in UCINET6. Results were visualized using NetDraw. Through frequency analysis, 7,542 keywords were extracted. Of them, 86 high-frequency keywords were selected for network analysis. The CONCOR analysis revealed seven key thematic clusters: school, public officials, military, family, anomie, suicide attempts, and suicide locations. This study contributes to a deeper understanding of interconnected socio-cultural factors influencing suicide dynamics in South Korea. By examining a large, diverse dataset over a ten-year period, this research offers new insights into the evolution of suicide-related discourse and the role of media in shaping public attitudes. Findings of this study provide valuable implications for suicide prevention strategies, policy-making, and future research on the role of media in shaping societal perceptions of suicide.