Database Reference
In-Depth Information
space of visualization techniques, as well as more focused exploration and anal-
ysis of particular data sets. Such flexibility in the overall process of visualization
almost always requires substantial expertise, typically requiring programming
skills. As a result, visualization software for exploration by experts often takes
the form of toolkits that are written in a popular programming language but
that encapsulate well-known visualization components and techniques.
One such toolkit, prefuse [44], provides a Java-based library of visualization
building blocks including views, visual encodings, processing algorithms, multi-
view coordinations, and a common data model that supports tables, trees, and
graphs. Graphs, hyperbolic trees, treemaps, and scatter plots support accessing,
filtering, rendering, and displaying data using a variety of layout and distortion
algorithms.
Similarly, the InfoVis Toolkit [34] is a set of Java visualization components
designed around OpenGL and a data model that represents tables, trees, graphs,
and metadata in column format for ecient selection, filtering, visual encod-
ing, and coordination. Views include scatter plots, parallel coordinate plots,
treemaps, and a variety of node-edge tree and graph displays that can incorpo-
rate fisheye lenses and dynamic labeling of items. Visualizations created in the
toolkit display textboxes, sliders, and other controls alongside views for dynamic
editing of visual encodings.
The extensible programming interfaces of both toolkits and those like them
provide a means to incorporate new components and techniques, in essence ex-
panding the scope of exploration, considered broadly, to include the results of
future visualization research.
Savvy Exploration and Analysis:
As described in the previous section, visu-
alization in the expert exploration category revolves more around programming
rather than around interaction in integrated user interfaces intended for design-
ing and building tools. Research on integrated visualization environments focuses
on packaging the exploratory capabilities of toolkits in ways that are accessible
to users who are visualization savvy but not necessarily visualization experts.
For instance, Improvise [102] is a self-contained Java application that appears
and behaves like other oce productivity applications based on the multiple doc-
ument desktop metaphor. Users build Improvise visualizations by interactively
constructing the data, queries, views, and coordinations of tools that can be
saved, opened, copied, and shared as self-contained Extensible Markup Language
(XML) documents. Users browse visualizations using the mouse and keyboard
to navigate and select data items in multiple coordinated views.
Similarly, GeoVISTA Studio [88] is an integrated visualization development
environment for building geovisualizations interactively using a graph-based vi-
sual coordination editor. Any component that conforms to the JavaBeans specifi-
cation can be a view. Development of new views by the community of GeoVISTA
Studio users has resulting in a large library of views utilized in numerous visual-
izations. A particular strength of GeoVISTA Studio is its extensive functionality
for representing and displaying geospatial information (based on the GeoTools
[57] open source Java GIS toolkit).
Search WWH ::
Custom Search