During my talk I will discuss my approach to visualizing large, multidimensional datasets. My results include research from a number of related areas like computer graphics, databases, and cognitive psychology. To date, I have investigated two aspects of the multidimensional visualization problem. First, I studied new methods for visually representing multidimensional data. These techniques address the problems of dataset size and data element dimensionality by exploiting the built-in processing of the human visual system. Second, I studied the effectiveness of a new database technique, knowledge discovery, for compressing and summarizing the important details buried in large datasets. I provided a direct comparison of four different algorithms, and showed how each algorithm can be extended and integrated into a visualization environment. My current work in visualization continues to study the use of color and texture for rapid, accurate, and effortless visual analysis.
I will begin with an overview of each of these areas of investigation. I will also show how results from each area combine to address the problem of multidimensional data visualization. I will provide a more detailed description of how we choose colors for use during visualization, to provide an example of how issues in graphics and cognitive psychology can impact research in scientific visualization. During all of this I will describe examples of how we have applied our theoretical results to real-world visualization problems. These include the analysis of sockeye salmon migration simulations, the display of abdominal aortic aneurisms, and the filtering and display of environmental conditions on topographical maps.