Mining Behavioural Patterns in Urban Mobility Sequences Using Foursquare Check-in Data from Tokyo

In a study of mobility and urban behaviour, we analyse a longitudinal mobility data set from a sequence mining perspective using a technique that discovers behavioural constraints in sequences of movements between venues. Our contribution is two-fold. First, we propose a methodology to convert aggregated mobility data into insightful patterns. Second, we discover distinctive behavioural patterns in the sequences relative to when in the day they were formed. We analyse sequences of venues as well as sequences of subcategories and categories to discover how people move through Tokyo. The results indicate that our methodology is capable of discovering meaningful behavioural patterns, that can be potentially used to improve urban mobility.