Rapid movement analytics for checkpoint data

Existing mobility research has tended to focus primarily on one important class of
movement data: trajectory data, such as GNSS trajectories. In contrast, checkpoint data, such as generated social media “check-ins” or smart card “swipes,” represents an ontologically distinct class of movement data. Checkpoint data exhibits fundamentally different spatial and temporal granularity characteristics when compared to trajectory data. Arguably underutilized and less well understood, this thesis investigates systematically the analytics of checkpoint movement data. The research approach develops a single, consistent conceptual model of checkpoint movement data, called the cordon network. The cordon network addresses one of the key challenges of using checkpoint data: the heterogeneity of checkpoint movement data sets. Unlike the trajectory data, which demands a primarily geometric framework, checkpoint data uses primarily graph data structures and algorithms. The cordon network is itself founded on the fundamental distinction between two complementary types of checkpoint movement data: transaction records and presence records. A set of primitive analytic operations are defined to work with both transaction and presence checkpoint data at varying scales. A software analytics toolkit that implements the cordon network model was developed, based on a graph database. To validate the implementation, a number of case studies were developed using real movement data sets. These example data sets exhibited highly heterogeneous characteristics, typical of checkpoint movement data. The results frame the commonalities of all checkpoint movement data, despite its apparent heterogeneity, including: the transaction/presence duality; distinct spatiotemporal granularity; and the intrinsic relationship between checkpoints their underlying movement environment. The cordon network provides a faithful representation of checkpoint movement data that successfully captures those features. The research outcomes provide a foundation and blueprint for future storage, sharing, and spatial analysis systems for checkpoint movement data.