Methods for Analysis of Spatio-Temporal Bluetooth Tracking Data

Analysis of people's movements represented by continuous sequences of spatio-temporal data tuples have received lots of attention in recent years. The focus of those studies was mostly GPS data recorded on a constant sample rate. However, the creation of intelligent location-aware models and environments also requires reliable localization in indoor environments as well as in mixed indoor/outdoor scenarios. In these cases, signal loss makes usage of GPS infeasible; therefore other recording technologies evolved. Our approach is analysis of episodic movement data. This data contains some uncertainties among time (continuity), space (accuracy), and the number of recorded objects (coverage). Prominent examples of episodic movement data are spatio-temporal activity logs, cell-based tracking data, and billing records. To give one detailed example, Bluetooth tracking monitors the presence of mobile phones and intercoms within a sensor's footprints. Usage of multiple sensors provides flows among the sensors. Most existing data mining algorithms use interpolation and therefore are infeasible for this kind of data. For example, speed and movement direction cannot be derived directly from episodic data; trajectories may not be depicted as a continuous line; and densities cannot be computed. Still, the data hold much information on group movement. Our approach is to aggregate movement in order to overcome the uncertainties. Deriving a number of objects for the spatio-temporal compartments and transitions among them gives interesting insights on the spatio-temporal behavior of moving objects. As a next step to support analysts, we propose clustering the spatio-temporal presence and flow situations. This work focuses as well on creation of a descriptive probability model for the movement based on Spatial Bayesian Networks. We present our methods on a real world data set collected during a football game in Nimes, France in June 2011.