Investigating physical encounters of individuals in urban metro systems with large-scale smart card data

Investigating physical encounters among individuals is important for various applications such as infectious disease modeling and friend recommendation. As enclosed spaces, public transit systems (e.g., buses and metros) in densely populated areas are locations where physical encounters occur numerously. Currently, encounter networks in bus systems have been investigated with the help of smart card data (SCD); however, no attempt has been made toward the metro systems, which is more challenging as the travel behaviors of metro passengers are complex but not recorded in the SCD in detail. This study proposed a novel framework for investigating physical encounters of individuals in urban metro systems with SCD. First, we developed a method to match passengers to specific trains, which can allow the segmentation of individual trips inside a metro system. Second, we proposed an approach to measuring the encounter frequencies and durations of each passenger pair by synthesizing their encounter behaviors in not only the train space, but also the entering/exiting space and the transfer space. Finally, using the SCD of Shenzhen, China, we analyzed the physical encounter patterns at a population scale, and demonstrated the potential of applying the encounter network to trace the spread of infectious diseases. Overall, this study provided a framework for evaluating physical encounters in metro systems with SCD, and revealed the underlying physical encounter patterns in the metro system of a metropolitan city, which is of considerable application value.