An Entropy-based Approach for Computing the Aesthetics of Interfaces

Abstract

The visual aesthetics of an interface profoundly affect user experiences, user brand preferences and the potential market viability of products and services. Although past computational approaches have been proposed based on multiple aesthetic dimensions to help designers quantify interface aesthetics, computational models have become very complex resulting in redundancy and overlap in some metrics. This paper introduces the entropy approach to simplify the modeling of graphic user interface aesthetics. By contrast with conventional linear models the proposed entropy approach is holistic and structured and as such it offers a more efficient approach to the quantification of visual aesthetics in interface design.