Having the ability to draw dynamic graphs is key to better understanding evolving relationships and analyzing the patterns and trends in a network. Traditional force-directed methods are not suitable for laying out dynamic graphs because of their design for static graphs. An alternative is to create an incremental version of the force multilevel multi-pole method (FM3); however, previous solutions are more susceptible to graph degradation, that is, graph illegibility due to long edges or edge crossings. This is typically caused when distant components are connected, resulting in long and overlapping edges. We present our incremental version of FM3 with a refinement scheme, which solves this problem by “refining” the parts of the graph with high energy. Our resulting visualization maintains readability of the graph structure and is efficient in laying out these changing networks. We evaluate the effectiveness of our method by comparing it with two previous online dynamic graph methods.
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)
- Computer Science Applications
- Geometry and Topology
- Computational Theory and Mathematics