An Empirical Study of Geographic User Activity Patterns in Foursquare

Abstract

We present a large-scale study of user behavior in Foursquare, conducted on a dataset of about 700 thousand users that spans a period of more than 100 days. We analyze user checkin dynamics, demonstrating how it reveals meaningful spatio-temporal patterns and offers the opportunity to study both user mobility and urban spaces. Our aim is to inform on how scientific researchers could utilise data generated in Location-based Social Networks to attain a deeper understanding of human mobility and how developers may take advantage of such systems to enhance applications such as recommender systems.