DG2CEP: A Density-Grid Stream Clustering Algorithm based on Complex Event Processing for Cluster Detection

Applications such as fleet and mobile task force management, or traffic control can largely benefit from the on-line detection of collective mobility patterns of vehicles, goods or persons. A common mobility pattern is a cluster, a concentration of mobile nodes in a certain region, e.g., a mass protest, a rock concert, or a traffic jam. Current approaches require previous knowledge of the locations where the cluster might happen. In this paper, we propose DG2CEP, an algorithm inspired by data mining algorithms and based on Complex Event Processing, for on-line detection of such clusters. It can detect the formation and dispersion of clusters from streams of position data without the need of specifying the possible locations of clusters in advance.