ℂ3-palette: Co-saliency based colorization for comparing categorical visualizations

Visual comparison within juxtaposed views is an essential part of interactive data analysis. In this paper, we propose a co-saliency model to characterize the most co-salient features among juxtaposed labeled data visualizations while maintaining class discrimination in the individual visualizations. Based on this model, we present a comparison-driven color design framework, enabling the automatic generation of colors that maximizes co-saliency among juxtaposed visualizations for better identifying items with the largest magnitude change between two data sets. We conducted two online controlled experiments to compare our colorizations of bar charts and scatterplots with results produced by existing single view-based color design methods. We further present an interactive system and conduct a case study to demonstrate the usefulness of our method for comparing juxtaposed line charts. The results show that our approach is able to generate high quality color palettes in support of visual comparisons of juxtaposed categorical visualizations.