Abstract: Abstract: To provide refined guidance for the optimization and balanced development of collaborative urban fire safety management, this paper establishes a Bayesian network for the effect of collaborative urban safety management based on data from four aspects of current urban fire safety source governance, fire risk prevention and control, safety supervision and management, and safety security capacity. Based on Bayesian network backward and forward reasoning, it is known that four aspects of the attainment values are identified, which should be strengthened from public fire protection infrastructure construction, emergency linkage construction, fire protection supervision and law enforcement force and fire safety training. And for the emergency management bureau and the fire and rescue division in the fire risk prevention and control of collaborative governance, it is necessary to strengthen the rescue forces allocation and emergency linkage construction.

Key words: Key words: fire safety, collaborative governance, Bayesian network, urban fire risk