Exploring hot spots at tourism destination by flow-based density clustering method

Advancements in mobile technology allows tourism researchers to access fine-grained location data reflecting the travelers’ flow at destinations. While travel flow data contains inclusive information about travelers’ mobility—including origins (i.e., where they depart) and destinations (i.e., where they visited)—existing studies grounded in central place theory focus predominantly on destination to identify tourism ‘hot spots’. This approach overlooks the dynamic spatial interactions, which provides limited understanding about travel mobility. Thus, this study aims (1) to propose an origin-to-destination flow-based density clustering (OD-FDC) algorithm taking into account the directional travel movement derived from central flow theory, and (2) to demonstrate the usefulness of the method by analyzing over 150,000 car navigation records from Jeju, South Korea. This study clearly delineates how the flow-based method, OD-FDC algorithm, models directional travel movement and identifies tourism hotspots by fully utilizing flow directions, intensities and spatial distributions. Results reveal that the OD-FDC algorithm outperforms point-based analysis method—traditional hotspot analysis (Getis-Ord Gi*) by uncovering dynamic spatial interactions. As a result, this research provides theoretical contributions to the literature on travel mobility and methodological implications in spatial analytic of flow data. The findings provide destination marketers with actionable insights in developing regional planning and marketing.