How to Deal with Different Densities of Urban Spatial Data? A Comparison of Clustering Approaches to Detect City Hotspots

In the field of urban data analysis, the detection of city hotspots is becoming a fundamental activity aimed at showing functions and roles played by each city area and providing valuable support for policymakers, scientists, and planners. However, since metropolitan cities are heavily characterized by variable densities, multi-density clustering algorithms might be more reliable than classic approaches to discover proper hotspots from urban data. This paper presents a study on hotspots detection in urban environments, by comparing two approaches, i.e., single-threshold and multi-density threshold ones, for clustering urban data. The experimental evaluation, carried out on a synthetic state-of-the-art multi-density dataset, shows that a multi-density approach achieves higher clustering quality than classic techniques.