An Adaptive Graphics Interface for Effective Visual Representation

Abstract

This paper presents a progression of experimental work aimed at incorporating principles of graphic presentation and preferences of users into a computer graphics system. These standard-s for effective visual representation of data are based on empirically developed criteria. Several different approaches for obtaining these criteria from picture attributes are described. First, relevant attributes are verified using a series of pictures created with a menu-driven program. Then representation criteria are developed by an automatic rule acquisition program in the form of condition-action rules. These rules can be utilized by an adaptive graphics system that can generate, change and refine images interactively. Practical applications are suggested.