Geography Reference
In-Depth Information
3.3.3 Challenges to create informative visualizations
The geographical and cartographic communities have many years of experience. Developers
know how to generate effective and clear map representations. Furthermore, the geovisual-
ization community has borrowed from this long heritage and have created many dynamic
and interactive exploratory visualization systems. Systems such as cdv (Dykes, 1997), Mon-
drian (Theus, 2002) and CommonGIS (Andrienko and Andrienko, 1999) include many
different forms, from scatter plots, bar charts and line graphs to parallel coordinate plots.
However, there is too much choice. Gahegan (1999) described one of the major challenges
with exploratory geovisualization as being the 'vast range of potential approaches and map-
pings', and the problem still exists today.
A user does not necessarily know which representation is most effective; consequently they
may miss out on some important information because of the representation methodology
they used. Thus, there is a need to examine which representation form is most effective for
a given task, and methods are required to aid the user in their exploration.
3.4 Interaction and manipulation
Both the developer and the user need to make choices, not only how to display the data,
but how to interact with the information. There are usually lots of parameters that can be
changed to alter the visual appearance of the display. In fact, just the simple act of altering
the colourmap can have a significant impact on what is emphasized and what is hidden in
the display.
From a developer's standpoint there are some important steps required to generate a
representation. There are different models to follow, but the principles remain the same.
First, a demonstration dataset is created from the original data. This is achieved through
filtering, aggregating and processing the original data. Second, the developer must choose
what visual form to use and then decide how to map the data into it. Often the data is first
mapped into an intermediate form, which is then rendered and displayed. The intermediate
form is a data structure that holds summary or pre-calculated information. For instance, if
the visualization is a three-dimensional surface, then triangles, coordinates, colour and other
geometric information may be stored. Other data may also be stored, such as cached values
to speed up rendering; associated arrays may be calculated to control coordinated views,
and a list of the user's selected elements. Third, the user then manipulates the parameters
to change how the information is displayed.
One of the initial choices of a developer and user is to decide where the result is dis-
played, that is, what alters when the parameters are changed. Parameter changes can be
made in four areas: first, to control the appearance of the demonstration data; second, to
control how the data is mapped to the display; third, to navigate the information such
as zooming, clipping or changing the level of abstraction of the information being dis-
played; and fourth, controlling the environment and other meta-information associated
with the system (such as determining where the views are placed on the screen, loading
or saving files or managing bookmarks or controlling other exploration management fa-
cilities). Finally, interaction can also be
coordinated
and thus affect multiple visualizations
simultaneously.
Search WWH ::
Custom Search