Trajectories and Their Representations - Mobility Data - page 17

behaviors. Well-known examples include the Meet , Convergence ,and Flock
behaviors that are discussed in Chapters 6 and 7 . Most spatio-temporal behav-
iors are generic behaviors, that is, they are supposed to be applicable to any
application domain. Semantic behaviors tend to be application-specific as they
handle the semantic aspects of trajectories and these semantic aspects are the
most frequently application-specific.
Behaviors are defined for trajectories. Still, many behaviors can also qualify
parts of trajectories. Most spatio-temporal behaviors that characterize the shape
of the trace of the trajectory can qualify whole trajectories as well as parts of
trajectories. For instance, the Straight behavior that characterizes a straight spa-
tial trace can be used for defining trajectories whose whole trace is straight as
well as trajectories that contain at least one straight segment whose length is
longer than some given threshold. Another example is the Flock behavior that
characterizes a group of trajectories that travel together (see Chapter 7 ). The
common travel may last during the whole trajectories or only during some part
defined by a time interval. On the other hand, behaviors that rely on some global
characteristic of the trajectories (e.g., some aggregation on the whole trajec-
tory, Begin and End) can apply only to whole trajectories. An example is the
StopMoreThanMove behavior that characterizes Stop/Move segmented trajec-
tories that spend more time during the stops than during the moves.
The number of behaviors that can be defined is unbounded, as any application
domain has its own typical requirements and any application adds its specific
requirements. We purposely abstain from trying to define a taxonomy of behav-
iors. Description of some works devoted to building such taxonomy can be
found in the Bibliographic Notes section. Of particular importance in clarifying
the broad vision of behaviors is the separation between individual and collective
behaviors, and the Sequence behaviors, both discussed in the next section.
1.5 Individual, Collective, and Sequence Trajectory Behavior
A very important feature of behaviors is whether they apply to single trajectories
or to groups of trajectories. The former are called individual behaviors, the latter
collective behaviors.
Definition 1.9. A trajectory individual behavior is a trajectory behavior that is
characterized by a predicate p ( T ) that bears on a single trajectory T .
Definition 1.10. A trajectory collective behavior is a trajectory behavior that is
characterized by a predicate p ( S ) that bears on a set of trajectories S .
The Tourist behavior we have seen in the previous section is an individ-
ual behavior: for each single trajectory in the data set we can decide whether