Presentation 2006-10-17
Clustering Movement Trajectory Data Based on Markov Chain Model
Yoshiharu ISHIKAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes a new approach to cluster and to summarize trajectories of a large number of moving objects. Former proposals on moving object clustering often use distances between trajectories for their clustering. However, since the distribution of movement patterns are not fully considered, the overall feature of the movements is not necessarily represented. In the proposed method, movement patterns are modeled according to the Markov chain model, and an information theory-based clustering is applied. The clustering method is an extension of the Information Bottleneck method and aims to find clusters that has minimum information loss (that means the increase of vagueness will be large). The proposed method relates the problem of summarization of trajectories and the increase of vagueness. This paper describe the basic idea of the proposed method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) moving objects / trajectories / clustering / Markov-chain model / Information Bottleneck method
Paper # DE2006-125,DC2006-32
Date of Issue

Conference Information
Committee DE
Conference Date 2006/10/10(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Vice Chair

Paper Information
Registration To Data Engineering (DE)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Clustering Movement Trajectory Data Based on Markov Chain Model
Sub Title (in English)
Keyword(1) moving objects
Keyword(2) trajectories
Keyword(3) clustering
Keyword(4) Markov-chain model
Keyword(5) Information Bottleneck method
1st Author's Name Yoshiharu ISHIKAWA
1st Author's Affiliation Information Technology Center, Nagoya University()
Date 2006-10-17
Paper # DE2006-125,DC2006-32
Volume (vol) vol.106
Number (no) 290
Page pp.pp.-
#Pages 6
Date of Issue