Uncovering Spatio‐Temporal Activity Patterns of Staying Behavior of Pedestrian Explorers Using Mobile Phone Navigation Trajectory

ABSTRACT Explorers, who constitute a distinct subset within the mobility group, hold significant importance in mobility research. Explorers use mobile navigation software to navigate unfamiliar territories, with walking navigation users demonstrating superior accuracy in pinpointing unfamiliar locations compared with other transportation modes. This study leveraged the substantial walking navigation trajectories of mobile phones to analyze the activity patterns of explorers. Trajectory data were initially employed to extract walking trajectories using the LightGBM model. Subsequently, geometric feature‐based methods were used to extract the stay points of the walking trajectories. To overcome the limitations of point‐of‐interest (POI) data, this study integrated area of interest (AOI) data. The activity semantics of stay points were discerned using the gravity model and Bayes' rules to analyze the spatiotemporal patterns in exploratory stay point activities. This study combines OpenStreetMap map data and questionnaire responses to analyze and validate the proposed methodology.