Dynamic Graph Visualization on Different Temporal Granularities

Dynamic graphs are typically represented in a time-to-space mapping with the goal to preserve the mental map in order to reduce cognitive efforts for comparison tasks. Such a mapping from time to space has the general drawback that space limitations are sooner reached than in corresponding time-to-time mappings to which graph animation belongs. Consequently, to get an overview about the dynamics in a graph sequence, space-efficient and compact visual encodings are used to show as many graphs in the sequence as possible. Temporal graph aggregation is hence a clever data transformation strategy, but negatively, it does not provide an overview about individual graphs nor does it show graph subsequences on finer time granularities. In this paper we describe a visualization technique that can visualize dynamic graphs in a time-to-space mapping and additionally, allows the graph analyst to interactively explore the dynamic graph data on different temporal granularities. Moreover, if the dynamic graph data is rather dense, it can be filtered by selecting density intervals. We illustrate the usefulness of our visualization tool by applying it to a dynamic graph dataset that simulates time-contiuously changing graphs.