Visual analytics of GPS tracks: From location to place to behaviour

The VAST 2014 Mini-challenge 2 [2] provided a set of GPS vehicle tracking and credit card transaction data with the aim of inferring behaviour of those using the vehicles and credit cards. In particular it required designing or applying a visual analytics system to identify typical behaviours with which atypical, suspicious activity could be contrasted. It is an example of the more general case of inferring behaviour from movement records (e.g., from mobile devices [6] or public bikeshare schemes [1]). In this brief paper we summarise the design taken to address the particular VAST challenge while proposing a general approach to movement-based behaviour detection.