Based on expert knowledge and data fusion methods, a Bayesian network model for accident severity influence factor analysis is constructed. The K2 algorithm and correlation analysis are used to learn the structure of the model, and the Bayesian method is used to learn the parameters of the model. The data base of road traffic serious accidents occurred from 2011-2018 is used to analyze the influence of human, vehicle, road and environment factors on accident severity. The results show that the accident severity does not increase with the increase of the total number of people carried by the vehicle, but is closely related to the driver's operation behavior and vehicle safety condition, and when the vehicle technical condition is poor or the driver's operation is improper, once the accident occurs, it is more likely to cause mass death and mass injury. The model can better identify the factors that have a significant impact on the severity of the accident. The results of the study can provide a basis for the industry management to have a deeper understanding of the predisposing factors of serious accidents and improve the safety level of the road traffic system.

road traffic;serious accidents;accident severity;Bayesian network