Reading urban land use through spatio-temporal and content analysis of geotagged Twitter data

Abstract

This study explores the possibilities of reading urban land use through geotagged social media data using temporal and content analysis. The advent of social media into the everyday life of cities has transformed the natural complexity of urban space. People's interaction with space and with the social context happens in a new hybrid space that is becoming a part of the reality of city life. The publicly shared content that people produce as a side product of their digital routine can be utilized for developing new analytical studies. Social media data is not merely a new method of analysis, but a window into the emerging urban processes. Hence, understanding the potential of social media data in urban studies could provide new tools for future urban planning. The current study investigates the legibility of urban land-use patterns through classifications of geotagged Twitter data, with the aim of exploring the degree of empirical viability of using social media data for urban design processes. With this aim in mind, the study proposes a framework for utilizing geotagged Twitter metadata. The framework is tested in a university campus in the city of Famagusta in Cyprus. First, the study establishes a data collection and filtering method. Second, data synthesis and classification of the data using GIS and Kernel Density Estimation is explained. Third, the paper explores possibilities for combining the content analysis and temporal analysis and aims to find the best fit for reading urban land use. The outcome shows promising results in reading urban land use through geotagged data.