On the Multi-Dimensional Acceleration of Stochastic Blockmodeling for Community Detection

Stochastic block partitioning (SBP) is a community detection algorithm that is highly accurate even on graphs with a complex community structure. However, SBP is much slower than more commonly used algorithms, such as Louvain, making SBP impractical for analyzing large real-world graphs with millions of edges. Thus, we aim to realize fast and accurate community detection on large graphs by accelerating the highly accurate SBP algorithm via sampling, parallel and distributed computing on a cluster as well as algorithmic optimization. We compare our approach to other community detection algorithms, showing that SBP accelerated with our methods on 64 compute nodes is up to 1,163x faster than the official "Graph Challenge" baseline SBP implementation, while still being more accurate than the Louvain and Leiden algorithms on large graphs.