HITPACER (Hierhical Trajectory Partitioning and Clustering for Mining Recurrent Travel Behavior)

This ongoing collaborative project focuses on mining recurrent travel patterns at variable spatio-temporal scale from mobility data that includes three major components: a) Refinement of the trajectories with incomplete or noisy observations, b) qualitative symbolic representation of refined trajectories enabling significant data compression and computational costs reduction while preserving key original features, c) hierarchical trajectory partitioning with subtrajectory clustering revealing recurrent travel behavior at multiple spatial and temporal scales.The project is conducted in collaboration with Lockheed Martin under support of National Geospatial Intelligence Agency