Information Technology Reference
In-Depth Information
overview). Loosely adhering to the structures given there, the movement mining
work included in this section is grouped into four subsections.
•
Segmentation and filtering
•
Similarity and clustering
•
Movement patterns
•
Exploratory analysis and visualization
Figure
3.3
gives an overview of movement mining tasks. The iconic examples are
embedded in the six space models accommodating the movement of point objects
introduced in Fig.
2.1
in Sect. 2. The following section exemplifies the development
and application of movement mining techniques through original research included
in this topic.
(a)
(b)
(c)
a
1
a
2
b
1.1
b
1.2
c
6
c
3
a
3
c
2
c
5
c
1
c
4
a
4
(d)
(e)
(f)
d
1
d
2
B
A
t
2
f
1
t
1
f
2
f
3
t
1
D
e
3
A
t
2
C
t
4
e
4
t
3
t
4
t
5
t
3
G
G
E
F
t
5
t
6
K
t
7
t
12
K
J
H
t
6
t
7
t
8
I
e
1
e
2
d
3
Fig. 3.3
Examples for movement mining tasks embedded in typical conceptual movement spaces.
a
Exploratory analysis and movement patterns,
home range
(
a
1
) and arrangement patterns
leadership
(
a
2
),
flock
(
a
3
),
single file
(
a
4
);
b
segmentation into segments expressing different sinuosity (
b
1
.
1
,
b
1
.
2
);
c
similarity and clustering: similar 3D shapes or origin-destination (from
H
to
K
vs. from
K
to
A
).
d
similarity and clustering:
d
1
and
d
2
show highly synchronous movement,
d
3
is an outlier;
e
sequence patterns: two trajectories both featuring a sequence
I
,
F
,
G
;
f
similarity and clustering
in a network space;
f
1
and
f
2
are more similar than
f
1
and
f
3
, all three trajectories build an origin-
destination cluster (from
A
to
K
)(
P9
. Laube
2009
) (Reprinted from
Behaviour Monitoring and
Interpretation, BMI, Smart Environments
, Gottfried, B. and Aghajan, H. (eds.), Laube, P., Progress
in Movement Pattern Analysis, p. 55, Copyright (2009), with permission from IOS Press)
