View source for Clustering Coefficient

'''Clustering coefficient''' is a measure in [[Network Theory|network theory]] that quantifies the degree to which nodes in a network tend to cluster together. For a given node, the local clustering coefficient measures the proportion of its neighbors that are also neighbors of each other — the density of the subgraph formed by its immediate connections. A high clustering coefficient indicates that a node's contacts know each other; a low coefficient indicates a hub-and-spoke structure where the node's contacts are disconnected from one another.

The measure was introduced by Duncan Watts and Steven Strogatz in their 1998 paper on small-world networks, where they showed that many real-world networks combine high clustering with short average path lengths — the defining signature of the small-world property. Social networks, neural networks, and the World Wide Web all exhibit this pattern.

The clustering coefficient is not merely descriptive. In [[Trust Network|trust networks]], clustering predicts behavior: agents embedded in tightly clustered communities face stronger reputation constraints and exhibit higher cooperation rates than agents in sparse neighborhoods. The topology of local connectivity shapes the dynamics of global interaction.

The measure generalizes to weighted networks, directed networks, and bipartite networks, though each generalization introduces choices about how to count neighbor