Understanding the Critical Causes of Safety Accident in Modular Construction Through Text Mining and Network Analysis Approaches
摘要
With the emphasis on green growth and sustainable development, modular construction (MC) has been increasingly adopted in both developed and developing countries. MC is considered a safe and efficient construction method since it shifts many aspects of on-site activities to off-site factories. However, the unique feature of MC can bring a new set of safety risks either in the manufacturing or construction stages. Existing studies investigate the safety accidents in MC through simple statistical analysis or expert interviews, but they fail to reveal the accident chain at a detailed activity level. This study aims to identify the critical causes of safety accidents in MC by using texting mining and network analysis approaches. First, 138 accident reports were collected from the Fatality and Catastrophe Investigation Summaries of the U.S. Occupational Safety and Health Administration and the corresponding government websites of China. Then, the network model of MC accident causes is established, and the indicators of the network model are calculated in order to highlight the critical causes. The findings of this study lay the foundation for the development of safety management systems for MC.