Where have you been? Using location clustering and context awareness to understand places of interest

Abstract

Mobile devices have access to multiple sources of location data, but at any particular time often only a fraction of the location information sources is available. Fusion of location information can provide reliable real-time location awareness on the mobile phone. In this paper we propose and evaluate a novel approach to detecting the places of interest based on density-based clustering. We address both extracting the information about relevant places from the combined location information, and detecting the visits to known places in the real time. In this paper we also propose and evaluate ContReMAR application - an application for mobile context and location awareness. We use Nokia MDC dataset to evaluate our findings, find the proper configuration of clustering algorithm and refine various aspects of place detection. © 2012 Springer-Verlag.