A method to compute the communicability of nodes through causal paths in temporal networks

TL;DR

This paper introduces a novel method to measure node communicability in temporal networks by analyzing causal paths, which helps identify influential nodes and predict epidemic spread dynamics.

Contribution

The paper presents a new approach that considers all causal paths in temporal networks and applies a damping procedure to identify high-communicability nodes.

Findings

High broadcast nodes accelerate epidemic spread when infected early.

High receive nodes can act as broadcasters when time is reversed.

The method accurately predicts epidemic dynamics in real-world networks.

Abstract

We present a method aimed to compute the communicability (broadcast and receive) of nodes through causal paths in temporal networks. The method considers all possible combinations of chronologically ordered products of adjacency matrices of the network snapshots and by means of a damping procedure favors the paths that have high communication efficiency. We apply the method to four real-world networks of face-to-face human contacts and identify the nodes with high communicability. The accuracy of the method is proved by studying the spread of an epidemic in the networks using the susceptible-infected-recovered model. We show that if a node with high broadcast is chosen as the origin of the outbreak of infection then the epidemic spreads early while it is delayed and inhibited if the origin of infection is a node with low broadcast. Receiving nodes can be treated as broadcasters if the arrow of time is reversed.