[Picture]
A multivariate visualization of CPU chips that varies 2D position plus hue, luminance, size, orientation, and transparency to encode five separate chip properties
A study of integrating perceptual guidelines from psychophysical experiments into an automated “visualization assistant.” Perceptual rules are converted into evaluation engines that critique a visualization and offer hints on how to improve it. Viewers answer a set of simple, application-independent queries about their dataset and analysis tasks. This is used in combination with the evaluation engines and mixed-initiative search algorithms from artificial intelligence to identify and improve promising visualization designs.