Information Technology Reference
In-Depth Information
tracking data with high precision, admittedly with a different notion of precision
( P19 .Bothetal. 2013 ).
Finally, a study reported on in Merki and Laube ( P16 . 2012 ) explicitly investigated
the influence of the choice of perspective when detecting interaction movement pat-
terns. Here, simulated movement of interacting animals and observed movement of
children in an outdoor game were modeled and tracked using both perspectives and
respective conceptual data models. The comparative experiments revealed that the
different conceptual models of space required different formalizations of the patterns
to be detected. For example, adhering to the Eulerian perspective resulted in loca-
tional information of MPOs being discretized: The only locational information about
an MPO at any time was the current edge between adjacent checkpoint nodes. The
interaction pattern investigated in this study thus required a formalized neighbor-
hood. For the Lagrangian case with its GPS fixes this neighborhood was modeled as
a disc with a given radius r . In the Eulerian case MPOs were formalized as neighbor-
ing when located on adjacent edges. These different formalizations not surprisingly
resulted in different analysis outcomes. Similarly, whereas the reaction time between
two objects meeting and one shying away proved crucial for the Lagrangian perspec-
tive, this criterion was less useful in the Eulerian perspective where again the given
edges dictated a coarse temporal granularity (see also the constraints issue discussed
in the next Sect. 2.2.2 ) .
2.2.2 Constraints to Movement
A second important characteristic refers to the degree of freedom moving objects
have in their movement. Whereas some objects can (seemingly) move wherever they
wish (e.g., flying birds), others are limited where and how they can move across
space (e.g., pedestrians in an urban street network).
The assumption of unconstrained and hence free movement in an Euclidean space
is popular as it allows for a very simple conceptual model of space. However, this sim-
plistic assumption may ignore crucial constraints of the moving objects and hence
inadequately model their movement. Consider for instance animals in movement
ecology. Some animals can't swim which turns waterways into insurmountable
barriers, while others will not cross steep mountain ranges or need waterways to
swim in ( P20 . Bleisch et al. 2014 ). Hence, their seemingly free movement capac-
ity is not free at all. Similarly, the homing pigeons studied in Laube et al. ( P3 .
2007 ) could indeed fly wherever they wanted, but the experimental setup was that
of a release-fly-back-to-loft scenario. Hence, movement azimuth distributions fol-
lowed to a certain degree the given release site-loft configuration. The freedom of
movement can also be a matter of scale (see also Sect. 2.3.2 ) . For instance, the
cross-scale movement analysis study in Laube and Purves ( P13 . 2011 ) revealed that
cows forage freely within their fenced paddock, but far reaching movement is lim-
ited by the fence. Whereas large scale foraging properties were not influenced by the
fence, edge effects became visible when investigating turning angles of the move-
 
Search WWH ::




Custom Search