An interactive web-based geovisual analytics platform for co-clustering spatio-temporal data

Xiaoling Wu, Ate Poorthuis, R. Zurita-Milla, M.J. Kraak

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Clustering methods are useful in analyzing patterns from big spatio-temporal data. However, previous studies typically rely on traditional clustering methods to explore spatial or temporal patterns. Co-clustering methods allow the concurrent analysis of spatial and temporal patterns by identifying location- and timestamp-clusters at the same time. By combining co-clustering with coordinated multiple views (CMV) in an interactive geovisual analytics platform, we facilitate the exploratory co-clustering analysis of spatio-temporal data and the results. Further enhanced by Web 2.0 standards, our geovisual analytics platform ease the access to co-clustering analysis from any web browser. More specifically, our platform allows users to upload data and to visually explore it using interactive CMV to help the selection of co-clustering parameters. Our platform also allows users to run co-clustering and to visually and interactively explore the results. To illustrate the use of our platform, we analyze Dutch annual average temperature for 28 stations from 1992 to 2011. Results show that our platform not only helps to get a better understanding of the dataset but also to choose the co-clustering parameters. Our platform helps to interpret the co-clustering results too, and it supports the extraction and exploration of complex patterns buried in the data. In the era of big data, our web-based platform enables the exploration of concurrent spatio-temporal patterns from large datasets by combing both computer power and human interpretative capabilities.
Original languageEnglish
Article number104420
Number of pages27
JournalComputers & geosciences
DOIs
Publication statusE-pub ahead of print/First online - 30 Jan 2020

Fingerprint

Web browsers
Temperature
Big data
method
analysis
temperature

Keywords

  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{c7f5a87d0805463c9a8fcbee65bb347d,
title = "An interactive web-based geovisual analytics platform for co-clustering spatio-temporal data",
abstract = "Clustering methods are useful in analyzing patterns from big spatio-temporal data. However, previous studies typically rely on traditional clustering methods to explore spatial or temporal patterns. Co-clustering methods allow the concurrent analysis of spatial and temporal patterns by identifying location- and timestamp-clusters at the same time. By combining co-clustering with coordinated multiple views (CMV) in an interactive geovisual analytics platform, we facilitate the exploratory co-clustering analysis of spatio-temporal data and the results. Further enhanced by Web 2.0 standards, our geovisual analytics platform ease the access to co-clustering analysis from any web browser. More specifically, our platform allows users to upload data and to visually explore it using interactive CMV to help the selection of co-clustering parameters. Our platform also allows users to run co-clustering and to visually and interactively explore the results. To illustrate the use of our platform, we analyze Dutch annual average temperature for 28 stations from 1992 to 2011. Results show that our platform not only helps to get a better understanding of the dataset but also to choose the co-clustering parameters. Our platform helps to interpret the co-clustering results too, and it supports the extraction and exploration of complex patterns buried in the data. In the era of big data, our web-based platform enables the exploration of concurrent spatio-temporal patterns from large datasets by combing both computer power and human interpretative capabilities.",
keywords = "ITC-ISI-JOURNAL-ARTICLE",
author = "Xiaoling Wu and Ate Poorthuis and R. Zurita-Milla and M.J. Kraak",
year = "2020",
month = "1",
day = "30",
doi = "10.1016/j.cageo.2020.104420",
language = "English",
journal = "Computers & geosciences",
issn = "0098-3004",
publisher = "Elsevier",

}

An interactive web-based geovisual analytics platform for co-clustering spatio-temporal data. / Wu, Xiaoling; Poorthuis, Ate; Zurita-Milla, R.; Kraak, M.J.

In: Computers & geosciences, 30.01.2020.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - An interactive web-based geovisual analytics platform for co-clustering spatio-temporal data

AU - Wu, Xiaoling

AU - Poorthuis, Ate

AU - Zurita-Milla, R.

AU - Kraak, M.J.

PY - 2020/1/30

Y1 - 2020/1/30

N2 - Clustering methods are useful in analyzing patterns from big spatio-temporal data. However, previous studies typically rely on traditional clustering methods to explore spatial or temporal patterns. Co-clustering methods allow the concurrent analysis of spatial and temporal patterns by identifying location- and timestamp-clusters at the same time. By combining co-clustering with coordinated multiple views (CMV) in an interactive geovisual analytics platform, we facilitate the exploratory co-clustering analysis of spatio-temporal data and the results. Further enhanced by Web 2.0 standards, our geovisual analytics platform ease the access to co-clustering analysis from any web browser. More specifically, our platform allows users to upload data and to visually explore it using interactive CMV to help the selection of co-clustering parameters. Our platform also allows users to run co-clustering and to visually and interactively explore the results. To illustrate the use of our platform, we analyze Dutch annual average temperature for 28 stations from 1992 to 2011. Results show that our platform not only helps to get a better understanding of the dataset but also to choose the co-clustering parameters. Our platform helps to interpret the co-clustering results too, and it supports the extraction and exploration of complex patterns buried in the data. In the era of big data, our web-based platform enables the exploration of concurrent spatio-temporal patterns from large datasets by combing both computer power and human interpretative capabilities.

AB - Clustering methods are useful in analyzing patterns from big spatio-temporal data. However, previous studies typically rely on traditional clustering methods to explore spatial or temporal patterns. Co-clustering methods allow the concurrent analysis of spatial and temporal patterns by identifying location- and timestamp-clusters at the same time. By combining co-clustering with coordinated multiple views (CMV) in an interactive geovisual analytics platform, we facilitate the exploratory co-clustering analysis of spatio-temporal data and the results. Further enhanced by Web 2.0 standards, our geovisual analytics platform ease the access to co-clustering analysis from any web browser. More specifically, our platform allows users to upload data and to visually explore it using interactive CMV to help the selection of co-clustering parameters. Our platform also allows users to run co-clustering and to visually and interactively explore the results. To illustrate the use of our platform, we analyze Dutch annual average temperature for 28 stations from 1992 to 2011. Results show that our platform not only helps to get a better understanding of the dataset but also to choose the co-clustering parameters. Our platform helps to interpret the co-clustering results too, and it supports the extraction and exploration of complex patterns buried in the data. In the era of big data, our web-based platform enables the exploration of concurrent spatio-temporal patterns from large datasets by combing both computer power and human interpretative capabilities.

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.cageo.2020.104420

U2 - 10.1016/j.cageo.2020.104420

DO - 10.1016/j.cageo.2020.104420

M3 - Article

JO - Computers & geosciences

JF - Computers & geosciences

SN - 0098-3004

M1 - 104420

ER -