Wang Feifei, Sun Zejun, Wang Guan, Sun Bohan
School of Information Engineering, Pingdingshan University, Pingdingshan, Henan, China.
School of Mechanical and Power Engineering, Zhengzhou University, Zhengzhou, Henan, China.
PLoS One. 2025 Aug 13;20(8):e0330199. doi: 10.1371/journal.pone.0330199. eCollection 2025.
In a complex network, the identification of node influence and the localization of key nodes play a crucial role in analyzing network structure and determining the positioning of nodes for information transmission control, resource redistribution, and network regulation. In this study, we propose a method for identifying influential nodes called "Multi-order Neighbors and Exclusive Neighborhood" (MNEN) after analyzing and investigating existing methods in the field. The MNEN method calculates a node's influence based on two factors: the node itself, its neighboring nodes, and its exclusive neighborhood. The influence of the node itself is determined by its degree value and K-shell (Ks) value, while the influence contribution of the neighbor node is calculated based on its degree value, Ks value, and the contribution from its exclusive neighbor node. To evaluate the algorithm's performance, we employ the SIR model as the benchmark and conduct simulation experiments to validate the MNEN method, comparing the results with other influential node identification methods. Our analysis demonstrates that the algorithm accurately identifies influential nodes in networks of different scales, yielding a positive overall impact and demonstrating a certain level of universality.
在复杂网络中，节点影响力的识别和关键节点的定位在分析网络结构以及确定用于信息传输控制、资源重新分配和网络调节的节点位置方面起着至关重要的作用。在本研究中，我们在分析和研究该领域现有方法之后，提出了一种名为“多阶邻居与排他邻域”（MNEN）的识别有影响力节点的方法。MNEN方法基于两个因素计算节点的影响力：节点本身、其相邻节点及其排他邻域。节点本身的影响力由其度值和K壳（Ks）值决定，而邻居节点的影响力贡献则基于其度值、Ks值以及其排他邻居节点的贡献来计算。为了评估该算法的性能，我们采用SIR模型作为基准并进行模拟实验以验证MNEN方法，将结果与其他有影响力节点识别方法进行比较。我们的分析表明，该算法能够准确识别不同规模网络中的有影响力节点，产生积极的总体影响并展现出一定程度的通用性。
