Urban Data and Spatial Segregation: Analysis of Food Services Clusters in St. Petersburg, Russia

Abstract

This paper presents an approach to study spatial segregation through clusterization of food services in St. Petersburg, Russia, based on analysis of geospatial and user-generated data from open sources. We consider a food service as an urban place with social and symbolic features and we track how popularity (number of reviews) and rating of food venues in Google maps correlate with formation of food venues clusters. We also analyze environmental parameters which correlate with clusterization of food services, such as functional load of the surrounding built environment and presence of public spaces. We observe that main predictors for food services clusters formation are shops, services and offices, while public spaces (parks and river embankments) do not draw food venues. Popular and highly rated food venues form clusters in historic city centre which collocate with existing creative spaces, unpopular and low rated food venues do not form clusters and are more widely spread in peripheral city areas.