Distributed Community Detection in Temporal Graphs

Community detection stands as a pivotal process in network analysis, having undergone extensive examination over the past 25 years within static network contexts.This task involves partitioning networks based on structural characteristics, specifically into classes of nodes exhibiting denser connections than the overall network.Moreover, in recent years, an additional layer of complexity has arisen, particularly in networks characterized by a static nature where each node and edge is assigned a valid time interval.Such historical graphs, unlike traditional graphs, incorporate a temporal dimension, allowing for the analysis of how connections between entities evolve and change over different time intervals.In this study, we present a distributed algorithm for static community detection within a query time interval in historical graphs.Specifically, when provided with a designated query time interval, our proposed method identifies communities by assessing the individual contributions of each edge and node within the graph during that specified time interval.To the best of our knowledge, this setting has not been considered before in the literature.