International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
In CAESA, we only need to go through the data set once in
order to calculate the parameters associated with the grid cells
at the bottom level, the overhead of CAESA is linearly
proportional to the number of objects with a small constant
factor.
3.0 r- STING *
2.5 | xe
8 20 s : s
$ 151 XUL
B 101 a CAESA
0.5 a A Lia se 0s le peril
0.0 3 ln | | +
0 2000 4000 6000 8000
Number of cells at bottom layer
Figure 7: Overhead comparison between STING and CAESA
when user change her/his focus
To obtain performance of CAESA, we implemented the house-
price example discussed in Section 3.2. We generated 2400 data
points(houses), the hierarchical structure has 6 layers in this test.
Figure 2 shows the expected result, Figure 3 shows the result of
our algorithm, and Figure 4 shows the result when user change
her/his focus according to the last query. From the experimental
result we can see apparently that our algorithm is valid and has
a high performance.
6. CONCLUSION
In this paper, we proposed a scalable and visualization-oriented
clustering algorithm for exploratory spatial analysis. It is of
high performance and low computing cost, and it can run focus-
changing clustering efficiently, can be adaptable to visualization
situation, that is, choosing appropriate relative clustering
granularity automatically according to current relative
visualization resolution. We built a prototype to demonstrate the
practicability of the proposed algorithm. Experimental results
show our algorithm is effective and efficient.
REFERENCES
Agrawal R., Gehrke J., Gunopulos D. et al., 1998. Automatic
subspace clustering of high dimensional data for data mining
applications. In Proc. of the ACM SIGMOD, pp. 94-105.
David Hand, Heikki Mannila and Padhraic Smyth, 2001.
Principles of Data Mining. Massachusetts Institute of
Technolology Press.
Ester,M., Kriegel, H. P., Sander,J. et al., 1996. A density-based
algorithm for discovering clusters in large spatial databases
with noise. In Proc.of 2nd Int. Conf. Knowledge Discovery and
Data Mining. pp. 226-231.
Guha S., Rastogi R. and Shim K., 1998. CURE: an efficient
clustering algorithm for large databases. In: Proc. of the ACM
SIG MOD, pp. 73784.
Halkidi, M., Batistakis, Y. and Vazirgiannis, M., 200].
Clustering algorithms and validity measures. In Proc of 13th
Intl. Conf. on Scientific and Statistical Database Management,
pp.3-22.
Hinneburg, A., Keim, D. and Wawryniuk, M., 2003. Using
projections to visually cluster high-dimensional data. /EEE
Computational Science and Engineering], 5(2), pp. 14 — 25.
Hu, B. and Marek-Sadowska, M., 2004. Fine Granularity
Clustering-Based Placement. /EEE Transactions on Computer-
Aided Design of Integrated Circuits and Systems. 23(4), pp. 527
- S36.
Jiawei Han, Micheline Kamber. Data Mining: Concepts and
Techniques. Morgan Kaufmann Publishers, 2001.
Matsushita, M. and Kato, T., 2001. Interactive visualization
method for exploratory data analysis. In Proc. of 15th Intl. Conf.
on Information Visualization, pp. 671—676.
Memarsadeghi, N. and O'Leary, D.P., 2003. Classified
information: the data clustering problem. /EEE Computational
Science and Engineering. 5(5), pp. 54 — 60.
Ng Ka Ka. E. and Ada Wai-chee Fu., 2002. Efficient algorithm
for projected clustering. In Proc.of ICDE, pp. 273—273.
Sheikholeslami G.. Chatterjee, S. and Zhang, A., 1998.
WaveCluster: a multi-resolution clustering app roach for very
large spatial databases. In Proc. of 24th VLDB, pp. 428 439.
Tin Kam Ho., 2002. Exploratory analysis of point proximity in
subspaces. In Proc. of 16th Intl. Conf. on Pattern Recognition.
Vol. 2, pp. 196 — 199
Wang Lian, Cheung, W.W., Mamoulis, N., et al., 2004. An
efficient and scalable algorithm for clustering XML documents
by structure. /EEE Transactions on Knowledge and Data
Engineering, 16(1), pp. 82 — 96
Wang, W., Yang, J. and Muntz, R., 1997. Sting: a statistical
information grid approach to spatial data mining. In Proc. of
VLDB, pp. 186-195.
Wang, W., Yang, J. and Muntz, R., 1999, Sting: An Approach
to Active Spatial Data Miniing. In Proc. of ICDE, pp. 116-125.
ACKNOWLEDGEMENTS
The work is supported by Hi-Tech Research and Development
Program of China under grant No. 2002AA 135340, and Open
Researches Fund Program of SKISE under grant No.
SKL(4)003, and IBM Research Award, and Open
Researches Fund Program of LIESMARS under grant No.
SKL(01)0303.
KEY
ABS
Geo;
effic
ina
urba
arch
prov
3Dt
navi
proc
spati
3D «
wall
effe
the €
f. T
The
com
simt
mod
incre
(WV
and
user
over
rece
visu
oper
tran:
info
(Lin
200(
mod
intei
deve
imp
and
(19€
cliet
3D:
inte:
the
web
of tl
virt