Analysis of Freeway Traffic Accidents Based on Association Rule Mining

To explore the characteristics and influencing factors of highway traffic accidents, this paper uses the association rule mining method to analyze the traffic accident data of some sections of G20 and G35 highways in Shandong Province. The results show that there is a strong correlation between traffic accidents and lane location and traffic signs. Rear-end collisions are prone to occur on road sections with multiple traffic signs, and it has a strong correlation with minor accidents. Bumping the guardrail is more likely to occur in the first lane of a road section with a single traffic sign and is more serious. These findings can help in choosing options to reduce the risk of traffic accidents on highways, reduce traffic casualties, and promote road safety in other areas.