A Survey To Location-Based Context Prediction

Abstract. In recent years, beginning with the Neural Network Home Project, several approaches addressing context prediction have been pub-lished. This paper gives an overview of the conducted research in the field of predicting location-based context information published during the last 14 years. The location of a user or an object offers the most important and common context information and is easy to collect us-ing modern smartphones. Therefore it is likely to be the most relevant and attractive context information new researches might be interested in. Research works discussed in this paper are evaluated with regard to aspects concerning to the data sets the authors used for the evaluation of their prediction approaches. Finally, a comparison of three state of the art context prediction approaches with three popular data mining techniques with respect to their prediction accuracy is presented. The approaches are applied to four different data sets containing location based context information