Detection and visualization of subspace cluster hierarchies

Elke Achtert*, Christian Böhm, Hans Peter Kriegel, Peer Kröger, Ina Müller-Gorman, Arthur Zimek

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Subspace clustering (also called projected clustering) addresses the problem that different sets of attributes may be relevant for different clusters in high dimensional feature spaces. In this paper, we propose the algorithm DiSH (Detecting Subspace cluster Hierarchies) that improves in the following points over existing approaches: First, DiSH can detect clusters in subspaces of significantly different dimensionality. Second, DiSH uncovers complex hierarchies of nested subspace clusters, i.e. clusters in lower-dimensional subspaces that are embedded within higher-dimensional subspace clusters. These hierarchies do not only consist of single inclusions, but may also exhibit multiple inclusions and thus, can only be modeled using graphs rather than trees. Third, DiSH is able to detect clusters of different size, shape, and density. Furthermore, we propose to visualize the complex hierarchies by means of an appropriate visualization model, the so-called subspace clustering graph, such that the relationships between the subspace clusters can be explored at a glance. Several comparative experiments show the performance and the effectivity of DiSH.

Original languageEnglish
Title of host publicationAdvances in Databases: Concepts, Systems and Applications : 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Proceedings
EditorsR. Kotagiri, P. R. Krishna, M. Mohania, E. Nantajeewarawat
PublisherSpringer
Publication dateDec 2007
Pages152-163
ISBN (Print)978-3-540-71702-7
ISBN (Electronic)978-3-540-71703-4
DOIs
Publication statusPublished - Dec 2007
Externally publishedYes
Event12th International Conference on Database Systems for Advanced Applications, DASFAA 2007 - Bangkok, Thailand
Duration: 9. Apr 200712. Apr 2007

Conference

Conference12th International Conference on Database Systems for Advanced Applications, DASFAA 2007
CountryThailand
CityBangkok
Period09/04/200712/04/2007
SeriesLecture Notes in Computer Science
Volume4443
ISSN0302-9743

Fingerprint

Visualization
Experiments

Cite this

Achtert, E., Böhm, C., Kriegel, H. P., Kröger, P., Müller-Gorman, I., & Zimek, A. (2007). Detection and visualization of subspace cluster hierarchies. In R. Kotagiri, P. R. Krishna, M. Mohania, & E. Nantajeewarawat (Eds.), Advances in Databases: Concepts, Systems and Applications: 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Proceedings (pp. 152-163). Springer. Lecture Notes in Computer Science, Vol.. 4443 https://doi.org/10.1007/978-3-540-71703-4_15
Achtert, Elke ; Böhm, Christian ; Kriegel, Hans Peter ; Kröger, Peer ; Müller-Gorman, Ina ; Zimek, Arthur. / Detection and visualization of subspace cluster hierarchies. Advances in Databases: Concepts, Systems and Applications: 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Proceedings. editor / R. Kotagiri ; P. R. Krishna ; M. Mohania ; E. Nantajeewarawat. Springer, 2007. pp. 152-163 (Lecture Notes in Computer Science, Vol. 4443).
@inproceedings{040d5172dad0436c90972a9123c0e70d,
title = "Detection and visualization of subspace cluster hierarchies",
abstract = "Subspace clustering (also called projected clustering) addresses the problem that different sets of attributes may be relevant for different clusters in high dimensional feature spaces. In this paper, we propose the algorithm DiSH (Detecting Subspace cluster Hierarchies) that improves in the following points over existing approaches: First, DiSH can detect clusters in subspaces of significantly different dimensionality. Second, DiSH uncovers complex hierarchies of nested subspace clusters, i.e. clusters in lower-dimensional subspaces that are embedded within higher-dimensional subspace clusters. These hierarchies do not only consist of single inclusions, but may also exhibit multiple inclusions and thus, can only be modeled using graphs rather than trees. Third, DiSH is able to detect clusters of different size, shape, and density. Furthermore, we propose to visualize the complex hierarchies by means of an appropriate visualization model, the so-called subspace clustering graph, such that the relationships between the subspace clusters can be explored at a glance. Several comparative experiments show the performance and the effectivity of DiSH.",
author = "Elke Achtert and Christian B{\"o}hm and Kriegel, {Hans Peter} and Peer Kr{\"o}ger and Ina M{\"u}ller-Gorman and Arthur Zimek",
year = "2007",
month = "12",
doi = "10.1007/978-3-540-71703-4_15",
language = "English",
isbn = "978-3-540-71702-7",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "152--163",
editor = "R. Kotagiri and Krishna, {P. R.} and M. Mohania and E. Nantajeewarawat",
booktitle = "Advances in Databases: Concepts, Systems and Applications",
address = "Germany",

}

Achtert, E, Böhm, C, Kriegel, HP, Kröger, P, Müller-Gorman, I & Zimek, A 2007, Detection and visualization of subspace cluster hierarchies. in R Kotagiri, PR Krishna, M Mohania & E Nantajeewarawat (eds), Advances in Databases: Concepts, Systems and Applications: 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Proceedings. Springer, Lecture Notes in Computer Science, vol. 4443, pp. 152-163, 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Bangkok, Thailand, 09/04/2007. https://doi.org/10.1007/978-3-540-71703-4_15

Detection and visualization of subspace cluster hierarchies. / Achtert, Elke; Böhm, Christian; Kriegel, Hans Peter; Kröger, Peer; Müller-Gorman, Ina; Zimek, Arthur.

Advances in Databases: Concepts, Systems and Applications: 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Proceedings. ed. / R. Kotagiri; P. R. Krishna; M. Mohania; E. Nantajeewarawat. Springer, 2007. p. 152-163 (Lecture Notes in Computer Science, Vol. 4443).

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

TY - GEN

T1 - Detection and visualization of subspace cluster hierarchies

AU - Achtert, Elke

AU - Böhm, Christian

AU - Kriegel, Hans Peter

AU - Kröger, Peer

AU - Müller-Gorman, Ina

AU - Zimek, Arthur

PY - 2007/12

Y1 - 2007/12

N2 - Subspace clustering (also called projected clustering) addresses the problem that different sets of attributes may be relevant for different clusters in high dimensional feature spaces. In this paper, we propose the algorithm DiSH (Detecting Subspace cluster Hierarchies) that improves in the following points over existing approaches: First, DiSH can detect clusters in subspaces of significantly different dimensionality. Second, DiSH uncovers complex hierarchies of nested subspace clusters, i.e. clusters in lower-dimensional subspaces that are embedded within higher-dimensional subspace clusters. These hierarchies do not only consist of single inclusions, but may also exhibit multiple inclusions and thus, can only be modeled using graphs rather than trees. Third, DiSH is able to detect clusters of different size, shape, and density. Furthermore, we propose to visualize the complex hierarchies by means of an appropriate visualization model, the so-called subspace clustering graph, such that the relationships between the subspace clusters can be explored at a glance. Several comparative experiments show the performance and the effectivity of DiSH.

AB - Subspace clustering (also called projected clustering) addresses the problem that different sets of attributes may be relevant for different clusters in high dimensional feature spaces. In this paper, we propose the algorithm DiSH (Detecting Subspace cluster Hierarchies) that improves in the following points over existing approaches: First, DiSH can detect clusters in subspaces of significantly different dimensionality. Second, DiSH uncovers complex hierarchies of nested subspace clusters, i.e. clusters in lower-dimensional subspaces that are embedded within higher-dimensional subspace clusters. These hierarchies do not only consist of single inclusions, but may also exhibit multiple inclusions and thus, can only be modeled using graphs rather than trees. Third, DiSH is able to detect clusters of different size, shape, and density. Furthermore, we propose to visualize the complex hierarchies by means of an appropriate visualization model, the so-called subspace clustering graph, such that the relationships between the subspace clusters can be explored at a glance. Several comparative experiments show the performance and the effectivity of DiSH.

U2 - 10.1007/978-3-540-71703-4_15

DO - 10.1007/978-3-540-71703-4_15

M3 - Article in proceedings

AN - SCOPUS:38049175016

SN - 978-3-540-71702-7

T3 - Lecture Notes in Computer Science

SP - 152

EP - 163

BT - Advances in Databases: Concepts, Systems and Applications

A2 - Kotagiri, R.

A2 - Krishna, P. R.

A2 - Mohania, M.

A2 - Nantajeewarawat, E.

PB - Springer

ER -

Achtert E, Böhm C, Kriegel HP, Kröger P, Müller-Gorman I, Zimek A. Detection and visualization of subspace cluster hierarchies. In Kotagiri R, Krishna PR, Mohania M, Nantajeewarawat E, editors, Advances in Databases: Concepts, Systems and Applications: 12th International Conference on Database Systems for Advanced Applications, DASFAA 2007, Proceedings. Springer. 2007. p. 152-163. (Lecture Notes in Computer Science, Vol. 4443). https://doi.org/10.1007/978-3-540-71703-4_15