One of the most studied problems in network science is the identification of those nodes that, once activated, maximize the fraction of nodes that are reached by a spreading process of interest. In parallel, scholars have introduced network effective distances as topological metrics to estimate the hitting time of diffusive spreading processes. Here, we connect the two problems – the influential spreaders identification and spreading processes’ hitting time estimation – by introducing a centrality metric, called ViralRank, which quantifies how close a node is, on average, to the other nodes in terms of the random-walk effective distance. We show that ViralRank significantly outperforms state-of-the-art centrality metrics in identifying influential spreaders for super-critical contact-network processes and for metapopulation global contagion processes. Our findings deepen our understanding of the influential spreaders identification problem, and reveal how reliable diffusion hitting-time estimates contribute to its solution.
