Maximizing spreading in complex networks with risk in node activation

TL;DR

This paper proposes a risk-aware method for selecting initial spreaders in complex networks, balancing influence and activation risk to maximize information dissemination.

Contribution

It introduces a novel risk-aware metric that considers activation costs and outbreak size, improving influence maximization strategies in real social networks.

Findings

Risk-aware metric outperforms benchmark centralities in simulations.

Optimal spreaders are not necessarily the highest-degree nodes.

Activation risk influences the selection of effective initial spreaders.

Abstract

It is widely acknowledged that the initial spreaders play an important role for the wide spreading of information in complex networks. Thus, a variety of centrality-based methods have been proposed to identify the most influential spreaders. However, most of the existing studies have overlooked the fact that in real social networks it is more costly and difficult to convince influential individuals to act as initial spreaders, resulting in a high risk in maximizing the spreading. In this paper, we address this problem on the basis of the assumption that large-degree nodes are activated with a higher risk than small-degree nodes. We aim to identify the effective initial spreaders to maximize spreading when considering both the activation risk and the outbreak size of initial spreaders. On random networks, the analytical analysis reveals that the degree of optimal initial spreaders does not correspond to the largest degree of nodes in the network but rather be determined by infection probability and difference of activation risk among nodes with different degree. Here, we propose a risk-aware metric to identify the effective spreaders on real networks. The numerical simulation shows that the risk-aware metric outperforms the existing benchmark centralities in maximizing the effective spreading.