Monitoring a Dynamic Contact Network With Online Community Detection

The tutorial guides readers through the process of monitoring a dynamic contact network using an online community detection algorithm called LabelRankT, which is implemented in MAGE’s graph analytics library. This algorithm efficiently detects communities in networks that change over time, which is crucial for identifying clusters that could facilitate rumor spreading. By creating a network from a dynamically generated dataset modeled with the Barabási–Albert graph generation method, the tutorial demonstrates how to track and update communities as contacts change, ensuring that clusters are accurately detected and analyzed. The utility developed in the tutorial not only identifies these clusters but also calculates the average size of each, offering insights into how quickly rumors might spread through the network. The tutorial also incorporates Memgraph's capabilities, showing how triggers can automate community updates with every graph change. As data streaming becomes more prevalent, such algorithms are increasingly important for handling large volumes of data and producing meaningful results.