Towards a taxonomy of movement patterns

Abstract

Areview of research that has been carried out on data mining and visual analysis of movement patterns suggests that there is little agreement on the relevant types of movement patterns and only few, isolated definitions of these exist. Since the research interest in this area has recently started to soar, we believe that this is a good time to approach the definition of movement patterns in a more systematic and comprehensive way. This paper intends to contribute to the development of a toolbox of data mining algorithms and visual analytic techniques for movement analysis by developing firstly a conceptual frame-work for movement behavior of different moving objects and secondly a comprehensive classification and review of movement patterns. We argue that this is indispensable as a basis for the development of pattern recognition and information visualization algorithms that are required to be efficient (i.e. usable on massive data sets), effective (i.e. capable of accurately detecting patterns not artifacts), and as generic as possible (i.e. potentially applicable to different types of movement data). We demonstrate the utilization of our classification by answering the question as to what extent eye tracking data can be seen as a proxy of other types of movement data. We have set up a moderated discussion platform in order to facilitate the further evolution of our proposed classification towards a consolidated taxonomy in a consensus process.