Exploring the spatial distribution of geo-tagged Twitter feeds via street-centrality measures

Abstract

The socio-spatial complexity of urban spaces is increasing with the broad reach of mobile devices and digital communication mediums. This research explores the quantifiable spatial distribution of urban activities through the lens of social media. Accordingly, two key objectives are addressed in this paper. First, testing the reliability of social media as a tool for understanding urban public activities; and second, exploring the effects of physical accessibility on the interactions of people with urban spaces. Furthermore, this study explores new methodological possibilities for reading urban space through social media and measurable accessibility. Three layers of data are used to address these objectives: a collection of geo-tagged public Twitter feeds, a geo-tagged name-generator survey, and the metric Euclidian centrality measures of the urban spatial network (closeness and betweenness). The findings show that the geo-tagged Twitter data can be a reliable tool for understanding the socio-spatial structure of urban public spaces. Results also suggest a variety of socio-spatial patterns arising from relating Twitter data to centrality measures with more emphasis on locality.