Scalability Considerations for Multivariate Graph Visualization

T. J. Jankun-Kelly , Tim Dwyer , Danny Holten (1) , Christophe Hurter (2) , Martin Nöllenburg (3) , Chris Weaver (4) , Kai Xu (5)

1 Department of mathematics and computing science [Eindhoven]
2 LII - ENAC - Equipe Informatique Interactive
3 Institute of Theoretical Informatics
4 Penn State - Pennsylvania State University
5 Center for Information Materials and Department of Material Science and Engineering

Real-world, multivariate datasets are frequently too large to show in their entirety on a visual display. Still, there are many techniques we can employ to show useful partial views-sufficient to support incremental exploration of large graph datasets. In this chapter, we first explore the cognitive and architectural limitations which restrict the amount of visual bandwidth available to multivariate graph visualization approaches. These limitations afford several design approaches, which we systematically explore. Finally, we survey systems and studies that exhibit these design strategies to mitigate these perceptual and architectural limitations.

Domaines: Interface homme-machine [cs.HC]

T. J. Jankun-Kelly, Tim Dwyer, Danny Holten, Christophe Hurter, Martin Nöllenburg, et al.. Scalability Considerations for Multivariate Graph Visualization. Multivariate Network Visualization : Dagstuhl Seminar #13201, Dagstuhl Castle, Germany, May 12-17, 2013, Revised Discussions, Part II, 8380, Springer, pp 207-235, 2013, Lecture Notes in Computer Science, 978-3-319-06792-6. ⟨10.1007/978-3-319-06793-3_10⟩. ⟨hal-01078325⟩