Quantifying Node Influence in Networks: Isolating-Betweenness Centrality for Improved Ranking

In complex networks, node impact refers to an individual node's significance or influence within the structure. The evaluation of the impact of the nodes in information transmission, prevention of pandemics, and resilience applications of the infrastructure is studied. Centrality measures are crucial for understanding the impact of particular nodes in the network structure. Most centrality measures, such as degree centrality, betweenness centrality, and eigenvector centrality, provide influential node information based on network aspects such as connection patterns, paths for communication, and influence propagation dynamics. However, these centrality measures could capture local and global information by balancing time complexity and spreading efficiency. This paper proposes an Isolating-Betweenness Centrality (ISBC) for quantifying node impact by incorporating the properties Betweenness Centrality and Isolating Centrality. The proposed measure evaluates a node's impact by considering local and global structural influence. We verify the SIR and IC epidemic models to evaluate ISBC's performance compared with conventional and recent centrality measures on real-world datasets. Furthermore, we show that the proposed measure exhibits improved spreading efficiency over recent and conventional measures with moderate time complexity.