Identification of Key Contributing Factors and Risk Propagation Paths in Safety Accidents at Chinese Chemical Enterprises

Applies text mining to 422 chemical-enterprise accident investigation reports, extracts 29 causative factors, and builds a complex accident-causation network with centrality, Bayesian sensitivity, and propagation-path analysis. It is relevant to Process Safety Management because it moves incident investigation beyond single-event root cause analysis toward systemic identification of recurring factors, factor interactions, and critical risk-control priorities.