Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Comp Stat

Visual matrix explorer for collaborative seriation

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

In this article, we present a web‐based open source tool to support cross‐disciplinary collaborative seriation with the following goals: to compare different matrix permutations, to discover patterns from the data, annotate it, and accumulate knowledge. Seriation is an unsupervised data mining technique that reorders objects into a sequence along a one‐dimensional continuum to make sense of the whole series. Clustering assigns objects to groups, whereas seriation assigns objects to a position within a sequence. Seriation has been applied to a variety of disciplines including archaeology and anthropology; cartography, graphics, and information visualization; sociology and sociometry; psychology and psychometrics; ecology; biology and bioinformatics; cellular manufacturing; and operations research. Interestingly, across those different disciplines, there are several commonly emerging similar structural patterns. Visual Matrix Explorer allows users to explore and link those patterns, share an online workplace and instantly transmit changes in the system to other users. WIREs Comp Stat 2012, 4:85–97. doi: 10.1002/wics.193

Figure 1.

Workflow from acquiring predicate data to visual knowledge mining.

[ Normal View | Magnified View ]
Figure 2.

A dataset rendered in different permutations.

[ Normal View | Magnified View ]
Figure 3.

An overview of the VME workspace.

[ Normal View | Magnified View ]
Figure 4.

Workspace with two highlighted patterns.

[ Normal View | Magnified View ]
Figure 5.

Viewing the contents of the pink region—the results of a drilldown.

[ Normal View | Magnified View ]
Figure 6.

An example repertory grid.

[ Normal View | Magnified View ]
Figure 7.

Repertory grid visualized in multiple permutations.

[ Normal View | Magnified View ]
Figure 8.

Two highlighted patterns: magenta for a concept of loveable woman and blue for a rejecting‐threatening person.

[ Normal View | Magnified View ]
Figure 9.

Violet and plum regions representing a concept of an ideal person.

[ Normal View | Magnified View ]
Figure 10.

Findings of mammal species by location, a subset.

[ Normal View | Magnified View ]
Figure 11.

A visualization of the fossils dataset, rows represent sites, columns mark species.

[ Normal View | Magnified View ]
Figure 12.

Drilldown of the black dense region. Art and zodiac permutations introduce clumps that were not detectable on the original matrix.

[ Normal View | Magnified View ]
Figure 13.

Two sites of insectivores and rodents in Germany and France.

[ Normal View | Magnified View ]
Figure 14.

Details of the comet‐tail violet pattern in the initial matrix.

[ Normal View | Magnified View ]

Related Articles

Brushing
JMP statistical discovery software
Coping with high dimensionality in massive datasets
Scientific Visualization

Browse by Topic

Data Mining > Clustering and Classification
Data Mining > Exploratory Data Analysis
Data Visualization > Visualization of High Dimensional Data
blog comments powered by Disqus

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts