Application of graph theory to mining the similarity of travel trajectories

Travel mobility has attracted considerable attention from tourism scholars. Studies have extensively discussed discovering key (i.e., collective) movement patterns. Recently, the advancement of information technology has allowed tourism researchers to obtain detailed information regarding travel digital footprints. This study, which analyzes mobile sensor big data, proposes a data mining approach to measure the similarity of travel trajectories by performing a pair comparison of individual trajectory. This method considers the spatial and temporal dimensions of travel flow to help identify trajectory similarity across individual travelers. Considering graph theory, this research also applies graph-based spatiotemporal analytics to identify important insights from complex travel mobility networks. As a result, this study suggests an innovative approach to assess travel trajectory similarity, which can be regarded as a type of data-driven clustering method. This paper also demonstrates the applicability of network science in travel mobility.