New Insights into Individual Activity Spaces using Crowd-Sourced Big Data

This article discusses ongoing work towards an accurate model of daily urban movements, calibrated in part by data from social media services. In particular, it investigates the connections between the language that is used in Twitter messages, the relative location of the user, and the significance of the location to an individuals’ geographical awareness, in order to elucidate possible activity patterns from spatial and textual social media data. The results show that there are words which are closely associated with home, others with the local community, and some with more remote locations including workplaces and cultural centres. By identifying important locations for individual users, and associating these with the words that are commonly used in such places, this research contributes to a better understanding of how spatially-attributed social media data can be used to derive useful intelligence about daily urban movement patterns.