Association Rule-Based Traffic Accident Impact Factors Analysis on Low-Grade Highways

Traffic safety is a wide concern that influences the residents’ life and property. Low-grade highways in China have characteristics like long mileage, more accidents, and low traffic safety levels. This paper investigates the influencing factors for traffic accidents on low-grade highways using association analysis. The Apriori algorithm was optimized to realize multidimensional and multilayer mining, then a model with three levels and 26 related attributes was built. Traffic accident data of low-grade highways in Suichang County, Lishui City, and Zhejiang Province were utilized to evaluate the proposed method. Finally, based on the analysis results of accident attributes and the mining results of association rules, the characteristics of low-grade highway traffic accidents and some attribute combinations that easily lead to accidents were summarized. The results show that the association analysis can be applied to investigate the influencing factors for traffic accidents.