Daisy Visualization for Graphs

Abstract

Since graphs are ubiquitous representations of data that are used in many applications, creating graph layouts is an important problem. These graph layouts are usefully discussed in terms of aesthetics that originated from mathematical concepts. In contrast, we explore the use of alternative aesthetics to inspire the visualization of graphs. We present Daisy Visualization, for which we have designed a new graph layout that is inspired by ornamental patterns of daisy flowers. In Daisy Visualization, graphs' attributes are mapped to floral elements to create an attractive information visualization that might more readily hold viewers' attention. As a practical use case we apply Daisy Visualization to the layout of ecological networks based on real ecosystem datasets. We show how specific attributes of ecological networks such as input/output edges, or respiration, can be mapped to floral elements. We conducted a qualitative assessment of Daisy Visualization, where we obtained overall positive feedback and interesting specific thoughts about various design decisions and possible future directions.