Reference: Chapt. 5 in Brodlie, etal, "Scientific Visualization: Techniques and Applications, Springer-Verlag, 1992
A data visualization system must be easy to use and not require sophisticated computer skills since many of the users will be domain specialists rather than computer scientists. It must be flexible to allow for application to many different types of data analysis problems. It must be extensible so that new analysis techniques can be easily added.
There are several different aspects to the user interface for a data visualization system. These include the following:
There are some issues in HCI that are unique or especially important in visualization systems. There will be a special emphasis on cognitive and perceptual aspects.
The reason for having visualization systems is to help people solve problems. Thus, the area of cognitive psychology, which is the study of how people solve problems, is highly relevant. It has been postulated that the semantics of the representation of a problem may determine how difficult the problem is to solve. Studies have shown that when people are given isomorphic problems (identical in form) but with different semantic representations, there is little carry over in the problem solution process. This means that what the user learns in solving the first problem, is not used in solving the second problem, even though they may be identical in form, but not in representation.
Visualization systems should be developed with a clear understanding of the users cognitive processes.
There are several perceptual issues that are important in data visualization systems. A distinction sometimes must be made between the issues involved in the display and communication of data and the issues involved in the investigative and understanding phase of visualization. Color is a major perceptual issue. For more on human perception in Scientific Visualization, go here.
Last modified on February 11, 1999, G.
Scott Owen, email@example.com
HyperVis Table of Contents