Meeting Pattern Detection from Trajectories in Road Network

As technological advancements in positioning devices progress, the analysis of spatio-temporal trajectory data has become increasingly critical, particularly in identifying the group movement pattern of moving objects. However, existing studies on group convergence behavior in urban networks require absolute temporal continuity, which may lead to the loss of interesting patterns. To overcome this challenge, we introduce a new pattern, the meeting pattern, which relaxes the time constraints on the convergence behavior. To effectively detect meeting patterns, we design and implement two algorithms, the tree structure-based MT-MPM algorithm and the ID partition-based IDP-MPM algorithm. Extensive experiments conducted on three datasets not only validate the time characteristics of the meeting pattern, but also confirm the effectiveness and efficiency of MT-MPM and IDP-MPM algorithms.