Abstract: In view of the uncertainty among forest fire risk factors and the influence of the intervention of emergency management and rescue forces on forest fire risk control, a set of forest fire risk assessment index system is constructed. Combined with the experience and knowledge of the field experts, based on the Bayesian Network model, the forest fire risk network map is drawn to clarify and deal with the uncertainty of forest fire. The weight of each index is calculated based on the information gain, and the contribution value of each index node in the network graph is determined. The weighted Bayesian network evaluation model is established. The forest fire risk assessment in Inner Mongolia Autonomous Region of China is carried out. The experiment shows that the model can effectively realize the risk assessment of forest fire and has a certain practical value.

Key words: entropy, information gain, Bayesian Network, forest fire, risk assessment