References: Community Structure in Networks¶
-
Community Structure in Networks - Wikipedia - Covers the definition of community structure, the resolution limit problem, benchmark graphs, and an overview of major detection algorithms including modularity-based and spectral methods.
-
Modularity (Networks) - Wikipedia - Explains the modularity quality function Q, its derivation from the null model (configuration model), and the known resolution limit that prevents detection of small communities in large graphs.
-
Louvain Method - Wikipedia - Describes the two-phase greedy modularity optimization algorithm, its O(n log n) scalability, and the refinements introduced in the Leiden algorithm to fix the disconnected-community problem.
-
Networks: An Introduction - Mark Newman - Oxford University Press - The definitive reference for network science; Chapter 11 covers community detection from first principles, including spectral methods, the Girvan-Newman edge-betweenness algorithm, and modularity maximization with rigorous derivations.
-
Mining of Massive Datasets - Jure Leskovec, Anand Rajaraman, Jeff Ullman - Cambridge University Press - Chapter 10 treats community detection at scale; covers the BigCLAM overlapping-community model, the conductance-based community scoring function, and practical graph-mining considerations for networks with hundreds of millions of edges.
-
Modularity and Community Structure in Networks (Girvan & Newman, 2004) - arXiv - The foundational paper introducing the modularity function Q and demonstrating that greedy modularity maximization recovers ground-truth communities in benchmark and real-world networks.
-
Fast Unfolding of Communities in Large Networks (Blondel et al., 2008) - arXiv - Introduces the Louvain algorithm; proves two-phase greedy optimization achieves near-optimal modularity in O(n log n) time and demonstrates results on networks with up to 100 million nodes.
-
Overlapping Community Detection at Scale: BigCLAM (Yang & Leskovec, 2013) - arXiv - Introduces the BigCLAM model, which assigns each node a real-valued membership vector per community; derives the log-likelihood objective and shows the model outperforms hard-partition methods on Amazon, YouTube, and DBLP.
-
Community Detection Algorithms — PyTorch Geometric Documentation - PyTorch Geometric Docs - Documents graph clustering utilities and community-related transforms available in PyG, including DMoN pooling and spectral clustering helpers that integrate directly into GNN training pipelines.
-
Papers With Code — Community Detection - Papers With Code - Aggregates state-of-the-art results, leaderboards, and linked code repositories for community detection benchmarks including SBM synthetic graphs, Amazon co-purchase, and the OGB node-clustering suite.