Scalable dynamic graph summarization

Large-scale dynamic graphs can be challenging to process and store, due to their size and the continuous change of communication patterns between nodes. In this work we address the problem of summarizing large-scale dynamic graphs, maintaining the evolution of their structure and the communication patterns. Our approach is based on grouping the nodes of the graph in supernodes according to their connectivity and communication patterns. The resulting summary graph preserves the information about the evolution of the graph within a time window. We propose two online, distributed, and tunable algorithms for summarizing this type of graphs. We apply our methods to several real-world and synthetic dynamic graphs, and we show that they scale well on the number of nodes and produce high-quality summaries.