Design and Application of an Attractiveness Index for Urban Hotspots Based on GPS Trajectory Data

Urban hotspots refer to regions where flourishing shopping centers are located, the travel volume is very large, and there is high traffic. The formation of hotspots is strongly correlated with many features, i.e., time, space, and the distribution of points of interest; however, most studies have used qualitative analyses to describe the relevance of these features, and there is a lack of quantitative analyses. Therefore, we propose the concept and a model of the attractiveness index of a hotspot that is used to quantify the spatio-temporal distribution of the hotspots and determine the degree of attractiveness to the residents. In addition, a novel algorithm of hotspot similarity is designed and implemented to improve the efficiency of summarizing the hotspot data, which is usually performed manually. We mine the data and determine different hotspots using a one-week GPS trajectory dataset collected from 6599 taxis in Kunming. Furthermore, we calculate the attractiveness index of hotspots and visualize their characteristics. The research results provide a scientific basis and reference for public infrastructure planning, land-value evaluation, store location planning, consumption recommendations, and other applications.