axis (in the perceptual space) is essential for searching and validating an interpretation. Yet MDS programs currently provide few visualization tools for such purposes. WebMDS makes use of the multimedia nature of WWW to graphically display the perceptual space. Such a visualization as several advantages over a paper-based visualization. First, as is the case for online surveys, the stimulus objects are no longer limited to be images. They can be music or video segments, as well as three-dimensional CAD models. Second, the designers can interact with the perceptual space. For instance, a region of the perceptual space can be enlarged to reveal finer details of the clusters in that region. Icons representing stimulus objects can have hyperlinks to more detailed information. A three-dimensional perceptual space can be rotated to expose occluded stimulus objects. Third, because clustering analysis and preference mapping are integrated into WebMDS, the clusters formed by stimulus objects and the bundles formed by ideal vectors can be concurrently visualized.
Depending on the dimensions of the perceptual space, different techniques are needed for visualization. WebMDS currently provides a 2-dimensional perceptual space visualization module for interactive display of a perceptual space using orthogonal axes, A snapshot of this module, which is implemented by using CGI programs and Java applets, is shown in Figure 2. In this visualization, the designers can interactively zoom in to any region, or click on any icon to see detailed information about the object. For visualizing perceptual spaces of four or higher dimensions, a multidimensional perceptual space visualization module using parallel coordinates (Inselberg 1995] has been developed. Figure 3 shows a 2-dimensional perceptual map for sporting event mascots. WebMDS also provides a 3-dimensional perceptual space visualization module using Virtual Reality Modeling Language (VRML), as illustrated in Figure 4.
1. Green, P. E., Carmone, F. J. Jr., and Smith, S. M: 1989, "Multidimensional Scaling: Concepts and Applications", Boston: Allyn and Bacon, 1989.
2. Inselberg, A.: 1985 "The Plane with Parallel Coordinates", Special Issues on Computational Geometry of The Visual Computer 1, 69-97.
3. Kruskal, J. B., and Wish, M.: 1978, "Multidimensional Scating", Beverly Hills, CA: Sage University Series.
4. Lin, Rung-Tai., private communications, 1995.
5. Moore, W. L., and Pessernier, E. A.: 1993, Product Planning and Management: Designing and Delivering Value, McGraw-Hill.
6. Schiffman, S. M., L. Reynolds, and F. W. Young, Introduction to Multidimensional Scaling, Academic Press, 1981.