Visualizing ordered bivariate data on node-link diagrams

• DOI: 10.1016/j.visinf.2023.06.003
• Publication Date: 2023-06-25T09:53:24Z
• Authors: Osman Akbulut; Lucy McLaughlin; Tong Xin; Matthew Forshaw; Nicolas S. Holliman
• Keywords: Node-link diagram; Quantitative evaluation; Edge visualization; Uncertainty visualization; 0202 electrical engineering, electronic engineering, information engineering; Information technology; 02 engineering and technology; Bivariate network visualization; T58.5-58.64

Node-link visual representation is a widely used tool that allows decision-makers to see details about network through the appropriate choice of metaphor. However, existing visualization methods are not always effective and efficient in representing bivariate graph-based data. This study proposes novel node-link model — entropy (Vizent) graph effectively represent both primary secondary values, such as uncertainty, on edges simultaneously. We performed two user studies demonstrate efficiency effectiveness our approach context static diagrams. In first experiment, we evaluated performance Vizent design determine if it equally well or better than alternatives terms response time accuracy. Three encodings use cues were selected from literature for comparison: Width-Lightness, Saturation-Transparency, Numerical values. compared various graphs ranging complexity 5 25 three different tasks. The participants achieved higher accuracy their responses using values; however, Width-Lightness Saturation-Transparency did show equal all Our results suggest increasing size has no impact was then values visualization. Wilcoxon signed-rank test revealed mean seconds significantly less when presented, while significant difference found. experiments encouraging believe justify good alternative traditional data