Street characteristics and human activities in commercial districts: A clustering-based approach application for Shenzhen

Abstract

As a significant public place, the commercial area has a potential correlation between its built environment and human activities. However, the current research primarily concentrates on the internal environment of the store and customer satisfaction, while the impact of some environmental features of the outer space of the business district on visitors is seldom systematically discussed. This study takes four commercial districts in Shenzhen as examples, and the streets were categorized into five types based on street characteristics using the cluster analysis method. The relationship between each type of street and the population distribution in the region was subsequently discussed. To this end, a holistic approach was adopted, integrating multi-source urban data such as street view panorama, points of interest (POI), and street and building vectors to describe the built environment. Furthermore, the distribution of people at different times, based on location-based services (LBS) data, was combined to establish statistical models of various streets in commercial districts and evaluate the relationship between street characteristics and human activities. The results demonstrate that the relationship between population distribution and spatial characteristics is different in the five types of streets. Different types of streets have their own advantages, and human activities in the business district are often not affected by this advantage, but by other characteristics. The impact of these factors varies significantly between weekdays and weekends. By systematically categorizing street types and assessing the impact of environmental factors on pedestrian flow, this study sheds new light on the renewal and development of urban commercial districts in the future.