Text Mining Analysis of Media Reports and Legal Sentiments on the Death Penalty System in South Korea

Purpose: This study aimed to analyze media reports and legal sentiments related to the death penalty system in South Korea by using text mining to extract topics from media articles and analyze sentiments in Naver posts. Methods: A total of 826 media articles and 1,051 Naver posts published from January 1, 2010, to March 31, 2024, were collected through Bigkinds and Textom. Text preprocessing, keyword filtering, network structure analysis, and topic modeling of the media articles were conducted using NetMiner, while sentiment analysis of the Naver posts was performed using Python. Results: Three key topics were identified through topic modeling: (1) the Ministry of Justice's death penalty sentencing for brutal criminals, (2) the stance of the National Assembly, National Human Rights Commission, and civic organizations on the abolition of the death penalty, and (3) the Constitutional Court decision on the death penalty: constitutionality or unconstitutionality. Sentiment analysis revealed that 74% supported retaining the death penalty, while 25% favored its abolition. Conclusion: This study differs from previous public opinion survey research in that it employs a text mining analysis method using big data to examine the overall public sentiment toward the legal system. This approach provides significant insights and can serve as foundational data for suggesting the future direction of the death penalty system in South Korea. The human rights of executioners should also be considered.