Modeling Systemic Risk Propagation in Public Assembly Fire Accidents: A Causal Network Approach Based on Apriori Analysis

Fire accidents in public assembly occupancies often arise from complex interactions among institutional failures, behavioral violations, and physical hazards. Traditional linear analyses cannot capture these dynamic and systemic causation mechanisms. Drawing on 185 official investigation reports of major fire incidents in China from 1998 to 2020, this study builds a Fire Accident Causation Complex Network Model (FACCNM). The model integrates the Apriori algorithm, complex network analysis, and the 24 Model to identify key contributing factors and high-risk causal chains in fire accidents. The findings show that "unauthorized design change" and "renovation with flammable materials" occur frequently and occupy structurally central positions in fire risk propagation. The highest-risk causal chain, "Absence of fire audit/acceptance → unauthorized design change", illustrates a typical pattern in which institutional oversight triggers behavioral and material vulnerabilities. The results offer actionable guidance for fire risk governance, such as prioritizing intervention points, conducting targeted safety audits, and designing proactive regulatory strategies. This study also helps shift fire safety management from reactive response to preventive control, especially in high-density public environments.