LOUVAIN METHOD

Greedy Modularity Maximization for Community Detection

24
1.0
0.20
1.0
The Louvain method (Blondel et al., 2008) maximizes modularity Q in two phases: first, greedily move each node to the neighbor community that increases Q most; second, aggregate communities into super-nodes and repeat. Modularity Q = (1/2m) Σ [A_ij − γ·k_i·k_j/2m] δ(c_i,c_j) where γ is the resolution parameter. Higher γ finds smaller, tighter communities. The algorithm scales to billions of nodes, making it the de facto standard for large-scale network analysis. Watch nodes reassign to color-coded communities in real time.