Spatial-Temporal Characteristics Analysis of Passenger Transfer Behavior: An Empirical Study Based on Smart Card Data in Shenzhen

Understanding passengers’ transfer behavior and analyzing their transfer characteristics are crucial for improving the service level of public transportation systems. This paper uses smart card data collected in Shenzhen, China, combined with GIS data, including bus and metro routes information, to analyze the spatial-temporal behavior characteristics of passengers transferring from metros to buses. Specifically, this paper uses the Latent Dirichlet Allocation (LDA) model, which allows for the interaction between spatial and temporal patterns, to simultaneously analyze passengers’ transfer behavior from both time and space dimensions. The findings reveal notable variances in these behaviors between weekdays and weekends, offering insights into passenger profiles. Moreover, the findings identify transfer bottlenecks at specific times and locations, which can inform targeted management measures to reduce resource waste. These insights are instrumental in refining the efficacy of public transport systems.