Normalized strength-degree centrality: identifying influential spreaders for weighted network

Influential spreaders are key nodes in networks that maximize or control the spreading processes. Many real-world systems are represented as weighted networks, and several indexing methods, such as weighted betweenness, closeness, k-shell decomposition, voterank, and mixed degree decomposition, among others, have been proposed to identify these influential nodes. However, these methods often face limitations such as high computational cost, non-monotonic rankings, and reliance on tunable parameters. To address these issues, this paper introduces a new tunable parameter-free method, Normalized Strength-Degree Centrality (nsd), which efficiently combines a node’s normalized degree and strength to measure its influence across various network structures. Experimental results on eleven real and synthetic weighted networks show that nsd outperforms the existing methods in accurately identifying influential spreaders, strongly correlating to the Weighted Susceptible-Infected-Recovered (WSIR) model. Additionally, nsd is a parameter-free method that does not require time-consuming preprocessing to estimate rankings.