Information Technology Reference
In-Depth Information
the application field of movement ecology. The framework is based on four inter-
acting mechanistic components of organismal movement: 1. the internal state (why
do animals move?), 2. motion capacity (how do they move?), 3. navigation capacity
(when and where to move?), and 4. the external factors affecting movement. The
primary goal of their conceptual framework is providing a basis for hypothesis gen-
eration and contributing to a better understanding of the causes of movement and its
role in ecological and evolutionary processes. Hence, whereas Nathan et al. (
2008
)
aims at understanding movement as a process in ecology, the work summarized in
this topic rather aims at conceptual models required for abstracting and representing
movement in spatio-temporal information systems.
Having said that, I agree with Nathan et al. (
2008
) that CMA requires a firm
integration of the movement paths and the embedding movement space. By contrast,
much related work in GIScience and computer science mainly focuses on the move-
ment path or trajectory alone. However, the following definitions illustrate the crucial
role of the underlying movement space, even though in most work this interrelation
is not explicitly discussed.
Work focusing on the shape and arrangement of movement traces is typically
based on some definition of a
trajectory
as a polyline in a 2D Euclidean space that
can self-intersect, built by a set of time stamped points (Gudmundsson et al.
2007
).
This geometry-oriented perspective is useful for geometry-based operations, as is
for example illustrated in Gudmundsson et al. (
2009
) using a variant of the Douglas-
Peucker path-simplification algorithm for compressing large volumes of movement
data (similar to the motivation in Richter et al. (
P18
.
2012
) but using a rather different
approach). The importance of the embedding movement space becomes obvious
when movement is modeled as the path an object moved along in a transportation
network. For example, Cao and Wolfson (
2005
) define a
road-snapped trajectory
as a set of visited edges in a 2D transportation network. Similarly, Kuijpers et al.
(
2010
) model network-based trajectories in a 3D space-time prism. In contrast to such
merely geometric definitions, for others a trajectory has a semantic loading. Early
work on temporal GIS for spatio-temporal reasoning about people's personal lifelines
anticipated the importance of semantic enrichment of event histories (Thériault et al.
2002
). In their piece “on a conceptual view of trajectories” Spaccapietra et al. (
2008
,
p.130) define a trajectory as a “user defined record of the evolution of the position
of an object that is moving in space during a given time interval in order to achieve
a given goal”. Some others go even further and require that raw movement data is
first cleaned and preprocessed, even interpolated and segmented before a trajectory
in a narrower sense is created (Yan et al.
2008
). The semantic perspective has also
produced a large amount of research on formal and qualitative modeling of movement
trajectories. For instance, (Noyon et al.
2005
) propose a
spatio-temporal trajectory
(STT) abstract data type and related operations suitable for representing and querying
semantic-based trajectories.
Whereas the last decade has seen a constant stream of work studying
Lagrangian
,
GPS-based movement data, recent years have seen the arrival of more and more
work benefitting from the increasing availability of
Eulerian
movement data (from
GSM mobile phone systems, RFID, WiFi, Bluetooth). Even though not explicitly
