Predicting triadic closure in networks using communicability distance functions

We propose a communication-driven mechanism for predicting triadic closure in complex networks. It is mathematically formulated on the basis of communicability distance functions that account for the “goodness” of communication between nodes in the network. We study 25 real-world networks and show that the proposed method predicts correctly 20% of triadic closures in these networks, in contrast to the 7.6% predicted by a random mechanism. We also show that the communication-driven method outperforms the random mechanism in explaining the clustering coefficient, average path length, average communicability, and degree heterogeneity of a network. The new method also displays some interesting features toward its use for optimizing communication and degree heterogeneity in networks.