Dynamic graph visualization techniques can be based on animated or static diagrams showing the evolution over time. In this paper, we apply the concept of small multiples representations to visually illustrate the dynamics of a graph. Node-link diagrams are used as the basic visual metaphor for displaying individual graphs of the sequence. To improve the readability of the diagram and reduce visual clutter we apply an edge splatting technique. Here, we discuss the benefits of splatted radial graph layouts on a modifiable 2D grid. Moreover, to obtain a more scalable dynamic graph visualization we interactively support a graph analyst by a Rapid Serial Visual Presentation (RSVP) feature to rapidly flip between the sequences of displayed graphs. The usefulness of the technique is illustrated in two case studies investigating a dynamic call graph and an evolving social network that consists of more than 1,000 graphs.
In this work, we extend existing dynamic graph visualization by three contributions:
Splatted radial layout: We generate a radial graph layout that uses one representative node for each vertex instead of two as in our previous work on edge splatting based on bipartite and radial bipartite layouts. This further reduces visual clutter by shortening the link lengths. Applying edge splatting to a radial layout makes dense graphs more readable compared to non-splatted links.
Modifiable grid for small multiples: In the traditional parallel edge splatting technique, we displayed the graph sequence in a 1D row. In this paper, we place the graph sequence on a modifiable 2D grid,
making the dynamic graph visualization more scalable in the time dimension.
Flip-book feature: To further improve the scalability in the time dimension we use a flip-book feature. This is an extension to the RSVP visualization in which a graph subsequence is shown in a comic striplike representation and then smoothly animated. The flip-book approach makes use of the perceptual abilities of the human visual system to rapidly recognize time-varying visual patterns.