스포츠 영역에서 나타나는 폭력의 유형 : 텍스트 마이닝을 적용한 판례분석

This study identified the causes of violence in the sports field through case law analysis using text mining. To achieve the purpose of the study, data were collected from 2010 to September 2021 using Lbox (https://lbox.kr/), which provides court judgments of all grades. The final 224 cases related to sports (sexual) violence were collected after removing duplicated data. In the data collection procedure, first, the search was conducted using keywords such as physical education, class, physical education teacher, coach, director, exercise, assault, violence, corporal punishment, molestation, sexual harassment, and sexual violence. Second, the data collection range was from 2010 to September 2021, when the judgment was pronounced. Third, case law documents were classified according to time and field. Fourth, the corpus delicti was extracted. The collected text data was preprocessed through konlpy and gensim libraries of Python 3. Keyword frequency analysis and topic modeling were performed, and similar words of each topic keyword were explored using the Word2vec algorithm. As the results of topic modeling and similar words analysis, topics were classified into four categories. Specifically, Topic 1 is 'School and Professional Sports Violence,' Topic 2 is 'Sexual Harassment and Sexual Violence in School Sports,' Topic 3 is 'Sexual Harassment and Sexual Assault in Academy Sports,' and Topic 4 is 'Measures and After-Effects of (Sexual) Violence.' The study can be used as basic information for establishing the violence system and policy and setting punishment standards in the sports field. It can also be used to set policy agendas and policy actors to decide when countermeasures against sports violence are urgently needed.