가스라이팅 관련 범죄기사 분석: 텍스트마이닝과 내용분석을 중심으로

Abstract

This study aims to analyze the main keywords and themes of criminal cases related to gaslighting through text mining techniques, and to reveal the characteristics of articles containing the concept of gaslighting through content analysis of individual case articles on gaslighting. To this end, 710 criminal case articles searched for 'gaslighting' as a topic in 11 major daily newspapers in Korea from 2019 to 2023 were analyzed using text mining techniques such as word clouding and topic modeling, and the contents of the cases were presented through content analysis. As a result of topic modeling these incidents, they were organized into four topics: 'Valley murders', 'Minor abuse and sexual crimes', 'Sexual assault by authority', and 'Sports and celebrity-related incidents', which were reported the most. By analyzing the content of articles centered around the same incidents, a total of 118 criminal cases were extracted. The main types of crimes were sexual crimes, accounting for 60%, followed by murder, assault, fraud, and others. Further, when categorizing the relationship between perpetrators and victims, it was found that crimes occurred mainly in workplace relationships, romantic relationships, family relationships (parents-children, spouses, siblings), and priestly relationships. This reaffirmed that such crimes utilizing psychological manipulation and domination occur in existing intimate relationships. Based on the analysis results, directions, and implications for future research on domestic gaslighting crimes were suggested.