A Data-Driven Approach to Trip Generation Modeling for Urban Residents and Non-local Travelers

摘要: Trip generation modeling is essential in transportation planning activities. Previous methods that depend on traditional data collection are inefficient and expensive. This paper proposed a novel data-driven trip method for urban residents non-local travelers utilizing location-based social network (LBSN) cellular phone conducted case study Nanjing, China. First, the point of interest (POI) LBSN were classified into various categories by service type, then, four features each category including number users, POIs, check-ins, photos aggregated traffic analysis zones to be used as explanatory variables models. We random tree regression select most important model inputs, models established based ordinary least square model. Then, an exploratory approach was test performance combination with identify best residents’ travelers’ functions. The results suggest land use compositions have significant impact generations, patterns different between travelers.