Analysis of the Spatio-temporal Distribution of Traffic Accidents Based on Urban Built Environment Attributes and Microblog Data

Previous studies have proven that urban built environments influence people’s travel behaviors, such as travel mode, travel routes, and travel frequency, which subsequently influences the spatio-temporal distribution of traffic flow and accidents. In addition, spatial heterogeneity means that the effects of urban built environments on accidents vary between regions. This chapter, therefore, uses a geographically weighted regression model and microblog data to analyze the spatio-temporal distribution of accidents from the perspective of urban built environments. In terms of spatial aspects, the results show that proximity to a subway can effectively reduce accidents, whereas proximity to primary schools, kindergartens, and hospitals is positively associated with accidents. In terms of temporal aspects, the number of accidents on weekdays is much higher than that on weekends. As the spatial effects of factors such as subways, schools, and hospitals on weekdays are quite different from those on weekends, transportation planning and management departments should devise traffic control plans that are based on the spatio-temporal effects of different variables. This study provides approaches for the government to use to adjust the attributes of urban built environments to reduce the number of traffic accidents.