Detecting Urban Local Areas Based on Configurational Properties
of Street Network:
A Case Study of Shenzhen
Keen Pan, Haofeng Wang
Abstract：This study explores to adapt the configurational analysis of space syntax into the network science community detection
algorithms. Using configurational properties (directional distance, angular choice, and reach measures) as weights, this
study takes the street network of Shenzhen as an example, and applies street character weighted community detection
methods for identifying local urban areas. Validation of the identified results is performed using per capita consumption
price data (e.g., Dianping platform records) combined with urban density indicators. By deciphering how urban local
areas are structured through street network connectivity, this research not only provides a novel theoretical perspective
to approach urban self-organization but also offers technical support for urban planning and design practices.
