Identifying Influential Spreaders in Complex Networks Using Neighborhood Network Structure

Abstract

Identifying influential spreaders is a hot topic in complex network research. While centrality-based algorithms are easy to implement, they often have lower accuracy. Topologybased algorithms are effective for identifying network center influential spreaders but may not perform well in identifying peripheral influential spreaders. To address these problems, we propose a Neighborhood Structure Centrality (NSC) algorithm, which utilizes structural embedding and clustering to collect various network structural information and calculates node influence based on both node and neighborhood structural information. We compare the NSC algorithm with twelve baseline algorithms on six public datasets and four synthetic network datasets and demonstrate its higher accuracy.