ABSTRACT
Data lies at the heart of modern wide-are monitoring systems. In this chapter, a framework for the analysis and characterization of spatiotemporal measured data is introduced within the realm of trajectory mining techniques. The notion of temporal distances is introduced, and techniques to measure distances are discussed. Drawing upon the notion of kernel-based dimensionality reduction techniques, methods to characterize the observed behavior in terms of spatial and temporal modes are introduced and tested on observational data.
