Geography Reference
In-Depth Information
of any other and the environment. These exploratory visualization (EV) environments
are highly interactive systems and rely on the premise that 'insight is formed through
interaction'.
If we, as developers and researchers, are to provide the best possible environment for the
users, we need to take stock of what functionality is being provided by current tools and
examine what we are doing well and what we are perhaps not doing so well. This chapter
is a systematic analysis of exploratory geovisualization using CMV techniques. We consider
the strengths and weaknesses of the area and make a comprehensive review of CMV for
geovisualization.
3.1.1 Exploration
The goal of exploration is to search, locate and find out something new. A user starting out
on the investigation process may not know anything about the data, let alone the questions to
ask. Thus the system should help the user to not only display visual results of the information
but also browse and locate pertinent information. Visual exploration, in particular, enables
the user to visually investigate the data. In short, the visual exploration enables the user to
try out some parameters, instantly view the results of the parameter change, manipulate
the data through selection and highlight operations and relate that information to other
sources and visualizations. The user may directly select some interesting elements using
a bounding box tool, see the selected list of results in both a scatter plot visualization
and a textual list of results, and perhaps edit the items on the selection list by adding
some more or deleting some in the list. Thus exploration is part of a larger discovery
process.
Various researchers have described this discovery process by a variety of models. In geo-
visualization specifically DiBiase (1990), focusing on the role of visualization in support of
earth science research, summarizes this discovery process as four stages. First, exploration
reveals the questions and achieves familiarity with the data through testing, experimenting,
acquiring the right skills and learning about the underlying model. Second, the user needs
to confirm the relationships that exist in the data though comparison operations, relating
the information to other explorations and disproving other hypotheses. Third, the results
are synthesized; this is achieved by identifying the pertinent features and summarizing the
content. Fourth, and finally, the results are presented and they are taught and demonstrated
at professional conferences and in scholarly publications.
DiBiase's model emphasizes the uncertainty and goal-seeking nature of the discovery pro-
cess; it is a conceptual model that encourages the user to personally explore the information,
thus to gain a better understanding of the information before presentation to a wider audi-
ence. Obviously, these traits are highly important in exploratory geographical visualization
and developers need to decide how these features map to individual tools. However, the
model de-emphasizes the processes of data preparation and simplifies how the results are
gathered and presented to the user.
Sense-making models, on the other hand, are more goal-oriented models and are often
discussed in the context of intelligence analysis. These models emphasize the whole process
from data preparation to hypothesis presentation (Thomas and Cook, 2005). Specifically,
sense-making according to Russell
et al.
(1993) 'is the process of searching for a representation
Search WWH ::
Custom Search