Network Centrality

Eigenvector, Betweenness & PageRank on random and scale-free graphs — node size encodes centrality
30
High centrality
Low centrality
Eigenvector Centrality
A node is important if connected to other important nodes. Computed as the dominant eigenvector of the adjacency matrix (power iteration). Google's original PageRank is a variant with a damping factor.