Escape velocity centrality: escape influence-based key nodes identification in complex networks

Evaluating and measuring key nodes in highly populated networks is essential to control the spreading effects of diseases or rumors. Although several incentive approaches have been proposed in complex networks to identify key nodes, these approaches still have many challenges. Most of the existing approaches consider just a single aspect of a node in a network. To cope with these challenges, based on the escape velocity formula, we propose Escape Velocity Centrality (EVC) approach to combine the concerned features of the network (i.e., local and global features) to measure key nodes in complex networks with spreading dynamics. Furthermore, we design an extended version of EVC (i.e., EVC + ) to enhance the overall performance. To evaluate the effectiveness of EVC & EVC + , we implemented the proposed model via real-world as well as artificial networks. The empirical results based on susceptible–infected–recovered (SIR) and Kendall's correlation evaluation modelshave demonstrated that EVC & EVC + outperformed the state-of-the-art centralities with remarkable margins of improvements with respect to all of the facets of evaluation.