Spatio-temporally constrained origin–destination inferring using public transit fare card data

With the wide use and accessibility of automated fare collection systems (AFC) data, public transit origin–destination (OD) inferring has facilitated various geo-spatial studies. Due to data errors and human travel activity dynamics, OD inferring is always challenging. Although a large number of researchers have proposed multiple OD inferring methods to extract the boarding and alighting information from the AFC data, challenges for alighting stop inferring remain. Very few researchers have considered the spatio-temporal constraints of human traveling with a time geography perspective. And omitting built environment attributes usually leads to less accuracy. Concerning these issues, this study proposes an alighting stop inferring method considering spatio-temporal, distance decay, and built environment constraints. We utilize two types of data sources for validating the method at the bus stop level and the zone level by proposing network-based methods. With these methods embedded into a bus OD inferring framework, we perform a case study using real public transit data and reveal the public transit traveling patterns over time and urban space. It proves that the framework can provide insights into public transit system amelioration and urban planning.