Handling Complex Multilayer Networks—An Approach Based on Visual Network Analytics

We present a visual analytics framework to aid the investigation of dynamic multilayer networks, focusing on the visual tracking of evolving communities. We were challenged by edges not persistent in time, rather hectic, inducing communities that are “volatile”. Working closely with field experts, three complementary perspectives are developed to tackle this problem: times series summarizing a community’s behavior over time; drag and drop a selected community assessing how a community splits in time; augmented Sankey diagrams to investigate the split and merge patterns among multiple communities. Coupled with expert knowledge, we efficiently support tracking communities emerging from media events.