Tracking Changes in Dynamic Social Networks using Event Logs and Network Snapshots

Event logs can be used to keep track of fine-grained changes in dynamic networks. An event log is a sequence of change events, where each event associates an action (add or remove) with an element (node or edge) and a timestamp. The main benefit of an event log is the ability to construct a static graph (snapshot) of the network at any time point within the lifetime of the dynamic network. We present algorithms for construction of event logs from temporal edge lists for several scenarios. We also present algorithms for construction of static graphs from event logs, eventually using a list of snapshots. We conduct experiments to demonstrate the utility of event logs, and to study the efficiency of static graph construction with respect to the number of used snapshots.