Trajectories and Their Representations - Mobility Data - page 4

These sequences may be very long, far longer than the ideal unit of processing
of the application. Often the processing unit is some segment of the movement
of the object instead of the whole movement itself. For instance, for animals'
study the segments may correspond to the daylight time; for employees of an
enterprise the segments are defined by working hours, for example, 8 A.M.-
6 P.M.; for hikers in a natural park segments may be defined as going from one
camp site to another camp site. These segments of movement are nowadays
called trajectories . They are the unit of interest in applications' processing of
movement data. They are the focus of this chapter.
While movement is inherently continuous, it cannot be captured as such in
computers where stored data is by definition discrete. The movement track that
stores movement data consists of a discrete sequence of records (transmitted
by the acquisition device or input by humans) containing the position in space
and time of the moving object. Movement tracks are application independent;
their precise format and content depend on the device. Movement tracks are
analyzed and transformed to produce application-dependent representations of
trajectories. Because applications can require very different representations of
trajectories (with differences in their structure as well as differences in their
content) we define in this chapter three main kinds of trajectory representations
that we identified as particularly significant and useful: continuous, discrete, and
segmented.
Yet trajectories are not the only way to represent movement. Other repre-
sentations have been designed to suit applications that need some global view
of movement, resulting from the aggregation of the data about movement of
individual moving objects. For example, movement can be represented as a
field of vectors within a given space perceived as a continuous field. The vec-
tors aggregate data from the individual tracks to represent, for a given instant,
some characteristics - usually speed and direction - of the movements at every
position in space. Similarly, applications willing to globally analyze the flow of
objects moving among a discrete set of points (e.g., popular places within a city)
will aggregate individual movement tracks into edges between nodes of a flow
network as described in Chapter 15 on network systems. Various representations
of aggregated movements in a continuous field are presented in Chapter 8 .In
this chapter we deal with trajectories only.
Furthermore, movement data is inherently uncertain, because of imprecise-
ness of the data sensing and data transmission devices, or because of human
inaccuracy and data entry errors if a position is manually acquired. This chap-
ter does not address this issue, but Chapter 5 discusses uncertainty issues and
approaches in detail.
Application users rarely reason about locations expressed as geographical
coordinates: “I am at the Eiffel Tower” is easier to understand than “I am at
48 ° 51 29 North, 2 ° 17 40 East.” To enable easier and richer use of movement