Developing a Spatial-Temporal Contextual and Semantic Trajectory Clustering Framework

Abstract: This paper reports on ongoing research investigating more expressive approaches to spatial-temporal trajectory clustering. Spatial-temporal data is increasingly becoming universal as a result of widespread use of GPS and mobile devices, which makes mining and predictive analyses based on trajectories a critical activity in many domains. Trajectory analysis methods based on clustering techniques heavily often rely on a similarity definition to properly provide insights. However, although trajectories are currently described in terms of its two dimensions (space and time), their representation is limited in that it is not expressive enough to capture, in a combined way, the structure of space and time as well as the contextual and semantic trajectory properties. Moreover, the massive amounts of available trajectory data make trajectory mining and analyses very challenging. In this paper, we briefly discuss (i) an improved trajectory representation that takes into consideration space-time structures, context and semantic properties of trajectories; (ii) new forms of relations between the dimensions of a pair of trajectories; and (iii) big data approaches that can be used to develop a novel spatial-temporal clustering framework.