Visual fingerprinting

A new visual mining approach for large-scale spatio-temporal evolving data

Jiansu Pu, Siyuan Liu, Huamin Qu, Lionel Ni

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Spatio-temporal data analysis has important applications in transportation management, urban planning and other fields. However, spatio-temporal data are often highly dimensional, overly large and contain spatial and temporal attributes, which pose special challenges for analysts. In this paper, we propose a new visual aided mining approach, Visual Fingerprinting (VF) for extremely large-scale spatio-temporal feature extraction and analysis. It adopts a visual analytics approach for spatio-temporal data analysis that can generate fingerprints for temporal exploration while preserving the spatial distribution in a region or on a road. Fingerprinting has been proposed to display temporal changes in spatial distributions; for example fingerprints for a region grid can well display temporal changes in the traffic situations of significant spots of a city. These fingerprints integrate important statistical and historical information related to traffic and can be conveniently embedded into urban maps. The sophisticated design of the visualization can better reveal frequent or periodic patterns for temporal attributes. We have tested our approach with real-life vehicle data collected from thousands of taxis and some interesting findings about traffic patterns have been obtained. The experiments validate our methods and demonstrate that our approach can be used for analyzing vehicle trajectories on road networks.

Original languageEnglish (US)
Title of host publicationAdvanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings
Pages502-515
Number of pages14
DOIs
StatePublished - Dec 1 2012
Event8th International Conference on Advanced Data Mining and Applications, ADMA 2012 - Nanjing, China
Duration: Dec 15 2012Dec 18 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7713 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other8th International Conference on Advanced Data Mining and Applications, ADMA 2012
CountryChina
CityNanjing
Period12/15/1212/18/12

Fingerprint

Spatio-temporal Data
Fingerprinting
Fingerprint
Spatial distribution
Mining
Traffic
Spatial Distribution
Urban planning
Data analysis
Attribute
Visual Analytics
Urban Planning
Feature extraction
Visualization
Road Network
Trajectories
Feature Extraction
Integrate
Trajectory
Grid

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Pu, J., Liu, S., Qu, H., & Ni, L. (2012). Visual fingerprinting: A new visual mining approach for large-scale spatio-temporal evolving data. In Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings (pp. 502-515). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7713 LNAI). https://doi.org/10.1007/978-3-642-35527-1_42
Pu, Jiansu ; Liu, Siyuan ; Qu, Huamin ; Ni, Lionel. / Visual fingerprinting : A new visual mining approach for large-scale spatio-temporal evolving data. Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings. 2012. pp. 502-515 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{4a95fb901afc4e07bffb8de5e257a19b,
title = "Visual fingerprinting: A new visual mining approach for large-scale spatio-temporal evolving data",
abstract = "Spatio-temporal data analysis has important applications in transportation management, urban planning and other fields. However, spatio-temporal data are often highly dimensional, overly large and contain spatial and temporal attributes, which pose special challenges for analysts. In this paper, we propose a new visual aided mining approach, Visual Fingerprinting (VF) for extremely large-scale spatio-temporal feature extraction and analysis. It adopts a visual analytics approach for spatio-temporal data analysis that can generate fingerprints for temporal exploration while preserving the spatial distribution in a region or on a road. Fingerprinting has been proposed to display temporal changes in spatial distributions; for example fingerprints for a region grid can well display temporal changes in the traffic situations of significant spots of a city. These fingerprints integrate important statistical and historical information related to traffic and can be conveniently embedded into urban maps. The sophisticated design of the visualization can better reveal frequent or periodic patterns for temporal attributes. We have tested our approach with real-life vehicle data collected from thousands of taxis and some interesting findings about traffic patterns have been obtained. The experiments validate our methods and demonstrate that our approach can be used for analyzing vehicle trajectories on road networks.",
author = "Jiansu Pu and Siyuan Liu and Huamin Qu and Lionel Ni",
year = "2012",
month = "12",
day = "1",
doi = "10.1007/978-3-642-35527-1_42",
language = "English (US)",
isbn = "9783642355264",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "502--515",
booktitle = "Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings",

}

Pu, J, Liu, S, Qu, H & Ni, L 2012, Visual fingerprinting: A new visual mining approach for large-scale spatio-temporal evolving data. in Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7713 LNAI, pp. 502-515, 8th International Conference on Advanced Data Mining and Applications, ADMA 2012, Nanjing, China, 12/15/12. https://doi.org/10.1007/978-3-642-35527-1_42

Visual fingerprinting : A new visual mining approach for large-scale spatio-temporal evolving data. / Pu, Jiansu; Liu, Siyuan; Qu, Huamin; Ni, Lionel.

Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings. 2012. p. 502-515 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7713 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Visual fingerprinting

T2 - A new visual mining approach for large-scale spatio-temporal evolving data

AU - Pu, Jiansu

AU - Liu, Siyuan

AU - Qu, Huamin

AU - Ni, Lionel

PY - 2012/12/1

Y1 - 2012/12/1

N2 - Spatio-temporal data analysis has important applications in transportation management, urban planning and other fields. However, spatio-temporal data are often highly dimensional, overly large and contain spatial and temporal attributes, which pose special challenges for analysts. In this paper, we propose a new visual aided mining approach, Visual Fingerprinting (VF) for extremely large-scale spatio-temporal feature extraction and analysis. It adopts a visual analytics approach for spatio-temporal data analysis that can generate fingerprints for temporal exploration while preserving the spatial distribution in a region or on a road. Fingerprinting has been proposed to display temporal changes in spatial distributions; for example fingerprints for a region grid can well display temporal changes in the traffic situations of significant spots of a city. These fingerprints integrate important statistical and historical information related to traffic and can be conveniently embedded into urban maps. The sophisticated design of the visualization can better reveal frequent or periodic patterns for temporal attributes. We have tested our approach with real-life vehicle data collected from thousands of taxis and some interesting findings about traffic patterns have been obtained. The experiments validate our methods and demonstrate that our approach can be used for analyzing vehicle trajectories on road networks.

AB - Spatio-temporal data analysis has important applications in transportation management, urban planning and other fields. However, spatio-temporal data are often highly dimensional, overly large and contain spatial and temporal attributes, which pose special challenges for analysts. In this paper, we propose a new visual aided mining approach, Visual Fingerprinting (VF) for extremely large-scale spatio-temporal feature extraction and analysis. It adopts a visual analytics approach for spatio-temporal data analysis that can generate fingerprints for temporal exploration while preserving the spatial distribution in a region or on a road. Fingerprinting has been proposed to display temporal changes in spatial distributions; for example fingerprints for a region grid can well display temporal changes in the traffic situations of significant spots of a city. These fingerprints integrate important statistical and historical information related to traffic and can be conveniently embedded into urban maps. The sophisticated design of the visualization can better reveal frequent or periodic patterns for temporal attributes. We have tested our approach with real-life vehicle data collected from thousands of taxis and some interesting findings about traffic patterns have been obtained. The experiments validate our methods and demonstrate that our approach can be used for analyzing vehicle trajectories on road networks.

UR - http://www.scopus.com/inward/record.url?scp=84872733598&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872733598&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-35527-1_42

DO - 10.1007/978-3-642-35527-1_42

M3 - Conference contribution

SN - 9783642355264

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 502

EP - 515

BT - Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings

ER -

Pu J, Liu S, Qu H, Ni L. Visual fingerprinting: A new visual mining approach for large-scale spatio-temporal evolving data. In Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings. 2012. p. 502-515. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-35527-1_42