건설 재해에 관한 연구동향의 의미연결망 분석

Semantic Network Analysis on the Research Trends of Construction Accident

According to the Ministry of Employment and Labor's industrial accident statistics over the past decade, the number of accident deaths in the construction filed is the highest among all industries. Although participants in the construction process and institutions like Ministry of Employment and Labor and Korea Occupational Safety and Health Agency for safety inspection, have been conducting intensive management over the past years to reduce deaths in the construction industry, there has been no significant reduction in occupational accident. The construction industry and academia have been conducting various research and development to improve construction site safety. However, because there is a limit to increasing safety at construction sites only through individual research, efforts should be made to understand between the studies. This study explores domestic research trends in studies related to construction accident using text mining techniques. For this, a corpus was compiled, comprising published papers related to 'construction accident' from KERIS and KISTI during a period from 2000 to 2020. From this corpus, keywords were extracted using KoNLPy in Python and then a Network was built for Semantic Network Analysis by using the keywords. As a result of this study, the keywords 'Worker', 'Accident' and 'Site' found to be the main Keywords as hubs. In addition, the analysis of communities within the network identified five research directions; direct accident factors such as types and original cause materials, specific site safety management factors, working environment factors due to social structural changes, human error, and safety climate in construction site. The results show the trend of research on topics related to construction accident in Korea, which can contribute to identifying domestic research directions and determining subsequent research topics in the field.