A Survey of Graph-Based Representations and Techniques for Scientific Visualization

Abstract

Graphs represent general node-link diagrams and have long been utilized in scientific visualization for data organization and management. However, using graphs as a visual representation and interface for navigating and exploring scientific data sets has a much shorter history yet the amount of work along this direction is clearly on the rise in recent years. In this paper, we take a holistic perspective and survey graph-based representations and techniques for scientific visualization. Specifically, we classify these representations and techniques into four categories, namely, partition-wise, relationship-wise, structure-wise, and provenance-wise. We survey related publications in each category, explaining the roles of graphs in related work and highlighting their similarities and differences. We also point out research trends and remaining challenges in graph-based representations and techniques for scientific visualization.