Measuring the diversity and dynamics of mobility patterns using smart card data

Currently, smart card data analytics has caused new insights of human mobility patterns. Many applications of smart card data analytics, which have been applied from the bus traffic operation optimization to the traffic network optimization. Although the human travel behavioral features have been observed and revealed based on these statistical data, the diversity and dynamics are fundamental features of mobility data, requiring an in-depth understanding of the dynamic temporal-spatial features of these patterns. This paper measures the diversity and dynamics of human mobility patterns based on the smart card data of Chongqing. First, from individual mobility patterns, the measurement results indicate that the mobility patterns of urban passengers are similar during weekdays, but there is a distinct difference between weekdays and weekends. Second, based on the aggregated mobility patterns, each station has its own temporal profile. Specifically, the profiles of some stations are similar, because the land use types around these stations are identical. Third, based on the complex network theory, stations are divided into different clusters in a temporal scale. Interestingly, though clusters of stations are changing over time, adjacent stations which with close ids are always in the same cluster, because these stations are close to each other in geography. The above findings can help policymakers to make appropriate scheduling strategies and improve the efficiency of public transportation.