While construction projects are in full swing, they are also accompanied by frequent safety accidents. The ultimate goal of safety management is to reduce accidents and avoid casualties. The marginal causes, core causes, and causal pathways of safety accidents are crucial for understanding the occurrence and development of accidents.
This study investigates 790 accident case reports in China from 2015 to 2022, utilizing text mining technology and Bayesian network structures to identify risk causal factors, risk causal networks, and risk transmission pathways.
The findings are as follows: (1) Although there are numerous causal factors, the core factors can be condensed into eight, and the causal sets are relatively independent. (2) Through Bayesian network reverse reasoning, eight unique risk transmission pathways for different types of accidents were derived, but “failure to fulfill safety management responsibilities according to the law” is the most critical causal factor in the risk transmission pathways. (3) The diversification of accident risk transmission pathways indicates that the focus of engineering safety management should vary.
The results of this study can deepen the understanding of the complex coupling and transmission relationships between risk factors and transmission relationships between risk factors, clarify the core causative factors of accidents and identify regulatory strategies for different types of accidents, which has certain guiding significance for improving the level of safety management at engineering sites.
