Trajectory clustering: a partition-and-group
framework
J. Lee, J. Han, and K. Whang. Proceedings of the 2007 ACM SIGMOD international
conference on Management of data, page 593--604. New York, NY, USA, ACM, (2007)
DOI: 10.1145/1247480.1247546
Abstract
Existing trajectory clustering algorithms group
similar trajectories as a whole, thus discovering
common trajectories. Our key observation is that
clustering trajectories as a whole could miss common
<i>sub</i>-trajectories. Discovering common
sub-trajectories is very useful in many applications,
especially if we have regions of special interest for
analysis. In this paper, we propose a new
<i>partition-and-group</i> framework for clustering
trajectories, which partitions a trajectory into a set
of line segments, and then, groups similar line
segments together into a cluster. The primary advantage
of this framework is to discover common
<i>sub</i>-trajectories from a trajectory database.
Based on this partition-and-group framework, we develop
a trajectory clustering algorithm <i>TRACLUS</i>. Our
algorithm consists of two phases: <i>partitioning</i>
and <i>grouping</i>. For the first phase, we present a
formal trajectory partitioning algorithm using the
minimum description length(MDL) principle. For the
second phase, we present a density-based line-segment
clustering algorithm. Experimental results demonstrate
that TRACLUS correctly discovers common
sub-trajectories from real trajectory data.
%0 Conference Paper
%1 lee-trajectory-clustering-partition-2007
%A Lee, Jae-Gil
%A Han, Jiawei
%A Whang, Kyu-Young
%B Proceedings of the 2007 ACM SIGMOD international
conference on Management of data
%C New York, NY, USA
%D 2007
%I ACM
%K clustering dynamic
%P 593--604
%R 10.1145/1247480.1247546
%T Trajectory clustering: a partition-and-group
framework
%U http://www.cs.uiuc.edu/~hanj/pdf/sigmod07_jglee.pdf
%X Existing trajectory clustering algorithms group
similar trajectories as a whole, thus discovering
common trajectories. Our key observation is that
clustering trajectories as a whole could miss common
<i>sub</i>-trajectories. Discovering common
sub-trajectories is very useful in many applications,
especially if we have regions of special interest for
analysis. In this paper, we propose a new
<i>partition-and-group</i> framework for clustering
trajectories, which partitions a trajectory into a set
of line segments, and then, groups similar line
segments together into a cluster. The primary advantage
of this framework is to discover common
<i>sub</i>-trajectories from a trajectory database.
Based on this partition-and-group framework, we develop
a trajectory clustering algorithm <i>TRACLUS</i>. Our
algorithm consists of two phases: <i>partitioning</i>
and <i>grouping</i>. For the first phase, we present a
formal trajectory partitioning algorithm using the
minimum description length(MDL) principle. For the
second phase, we present a density-based line-segment
clustering algorithm. Experimental results demonstrate
that TRACLUS correctly discovers common
sub-trajectories from real trajectory data.
%@ 978-1-59593-686-8
@inproceedings{lee-trajectory-clustering-partition-2007,
abstract = {Existing trajectory clustering algorithms group
similar trajectories as a whole, thus discovering
common trajectories. Our key observation is that
clustering trajectories as a whole could miss common
<i>sub</i>-trajectories. Discovering common
sub-trajectories is very useful in many applications,
especially if we have regions of special interest for
analysis. In this paper, we propose a new
<i>partition-and-group</i> framework for clustering
trajectories, which partitions a trajectory into a set
of line segments, and then, groups similar line
segments together into a cluster. The primary advantage
of this framework is to discover common
<i>sub</i>-trajectories from a trajectory database.
Based on this partition-and-group framework, we develop
a trajectory clustering algorithm <i>TRACLUS</i>. Our
algorithm consists of two phases: <i>partitioning</i>
and <i>grouping</i>. For the first phase, we present a
formal trajectory partitioning algorithm using the
minimum description length(MDL) principle. For the
second phase, we present a density-based line-segment
clustering algorithm. Experimental results demonstrate
that TRACLUS correctly discovers common
sub-trajectories from real trajectory data.},
acmid = {1247546},
added-at = {2011-11-02T17:57:15.000+0100},
address = {New York, NY, USA},
author = {Lee, Jae-Gil and Han, Jiawei and Whang, Kyu-Young},
biburl = {https://www.bibsonomy.org/bibtex/241a7b6a683ad0be44b78114cac647153/mhwombat},
booktitle = {Proceedings of the 2007 ACM SIGMOD international
conference on Management of data},
description = {Trajectory clustering},
doi = {10.1145/1247480.1247546},
interhash = {fee78bf6593e52756fa1ead6cec0a2e6},
intrahash = {41a7b6a683ad0be44b78114cac647153},
isbn = {978-1-59593-686-8},
keywords = {clustering dynamic},
location = {Beijing, China},
numpages = {12},
pages = {593--604},
publisher = {ACM},
series = {SIGMOD '07},
timestamp = {2016-07-12T19:25:30.000+0200},
title = {Trajectory clustering: a partition-and-group
framework},
url = {http://www.cs.uiuc.edu/~hanj/pdf/sigmod07_jglee.pdf},
year = 2007
}