Information Technology Reference
In-Depth Information
(a)
(b)
H
3
H1
H2
H3
H4
20
H2
H1
15
H1
H
4
10
H2
H3
5
H4
time (hour)
0
0
50
100
150
200
250
300
(c)
(d)
H
H
H
H4
H1 H2 H3 H4
0 0.282 0.552 0.768
0.282 0 0.534 0.760
0.552 0.534 0 0.864
0.768 0.760 0.864 0
H1
H2
H3
H4
5
10
15
20
25
30
35
40
45
50
point index
PHH
PHL
PLH
PLL
MMM
NLH
NLL
NHH
NHL
Fig. 3.6
Trajectory similarity. Computation of normalized weighted edit distance (NWED) for
four hurricane trajectories.
a
spatial footprint of hurricane trajectories;
b
hurricane speed profiles;
c
segmented speed profiles;
d
NWEDpair-wise distancematrix (
P14
. Dodge et al.
2012
) (Republished
from Dodge, S., Laube, P., and Weibel, R., Movement Similarity Assessment Using Symbolic
Representation of Trajectories.
International Journal of Geographical Information Science
, 26(9),
pp. 1563-1588, 2012, Taylor & Francis, DOI:10.1080/13658816.2011.630003)
3.2.3 Movement Patterns
Mining movement patterns is the quintessence of movement mining. This topic fea-
tures work on coordination patterns such as
leadership
in Andersson et al. (
P5
.
2008
)
and
flocking
2
in Laube et al. (
P6
.
2008b
) and Laube et al. (
P12
.
2011a
), and Both et al.
(
P19
.
2013
) as well as reaction patterns such as
pursuit and escape
,
confrontation
,
or
avoidance
in Merki and Laube (
P16
.
2012
). Examples for movement patterns are
illustrated for
leadership
in Fig.
3.2
and
pursuit and escape
in Fig.
3.7
. This section
revisits the work on movement patterns included in this topic, reviews strengths and
weaknesses of the included work and thereon presents three suggestions for good
practice in movement mining.
3.2.3.1 Definitions Grounded in Application Theory
Definitions of movement patterns should be grounded in the theory of the respec-
tive application domain. Andersson et al. (
P15
.
2008
) base the conceptualization
of the pattern
leadership
on detailed descriptions of the involved processes and
2
Note, the research on
flocking
featured in this topic combines data mining concepts with decen-
tralized spatial computing principles. This chapter focuses on the data general data mining aspects,
Chap.
4
on the specifics of mining movement patterns in a decentralized setting.
