텍스트 마이닝을 이용한 산업재해 원인의 탐색적 연구

Abstract

This study conducted network analysis by utilizing text mining among keywords extracted from the causes of industrial accidents. To address this, Textom was utilized to examine the causes of industrial accidents covered in the abstracts of research papers on ‘risk assessment’, and attempted to search their correlation. Keywords including “chemicals”, “fire”, and “explosion” had a high frequencies in occurrence, and many studies on risk assessment have been conducted around “chemicals”. Moreover, it is necessary that often cause industrial accidents are approached with risk awareness. After the CONCOR analysis, four clusters were formed which will help identify hazards and risk factors in risk assessments, adequately analyze assessment contents, and decide risk assessments.