File(s) under permanent embargo
Efficiently retrieving longest common route patterns of moving objects by summarizing turning regions
conference contribution
posted on 2011-06-08, 00:00 authored by Guangyan HuangGuangyan Huang, Y Zhang, J He, Z DingThe popularity of online location services provides opportunities to discover useful knowledge from trajectories of moving objects. This paper addresses the problem of mining longest common route (LCR) patterns. As a trajectory of a moving object is generally represented by a sequence of discrete locations sampled with an interval, the different trajectory instances along the same route may be denoted by different sequences of points (location, timestamp). Thus, the most challenging task in the mining process is to abstract trajectories by the right points. We propose a novel mining algorithm for LCR patterns based on turning regions (LCRTurning), which discovers a sequence of turning regions to abstract a trajectory and then maps the problem into the traditional problem of mining longest common subsequences (LCS). Effectiveness of LCRTurning algorithm is validated by an experimental study based on various sizes of simulated moving objects datasets. © 2011 Springer-Verlag.
History
Event
Pacific-Asia Conference on Knowledge Discovery and Data Mining (15th : 2011 : Shenzheng, China)Volume
6634Issue
Part 1Series
Lecture Notes in Artificial IntelligencePagination
375 - 386Publisher
SpringerLocation
Shenzheng, ChinaPlace of publication
Berlin, GermanyPublisher DOI
Start date
2011-05-24End date
2011-05-27ISSN
0302-9743eISSN
1611-3349ISBN-13
9783642208416Language
engPublication classification
E Conference publication; E1.1 Full written paper - refereedCopyright notice
2011, SpringerEditor/Contributor(s)
J Huang, L Cao, J SrivastavaTitle of proceedings
Advances in Knowledge Discovery and Data MiningUsage metrics
Categories
No categories selectedLicence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC