Influential nodes identification based on hierarchical structure

Identifying influential nodes is an important research topic in complex network analysis, with significant implications for understanding and controlling propagation processes. While extant methods for assessing node influence rely heavily on network topology, often overlooking the dynamic interactions and propagation patterns within networks. In this paper, we propose the Hierarchical Structure Influence (HSI) method. The HSI method evaluates the potential outbreak size of nodes by modeling their infection sequences and paths according to a network’s hierarchical structure, and integrating propagation probabilities to estimate these outbreak sizes accurately. It accounts for infections occurring across different node layers, intra-layer, and heterogeneous infection routes of varying lengths. To validate its effectiveness, HSI is compared with seven state-of-the-art methods across nine real-world networks. Experimental results reveal that HSI outperforms other methods in terms of ranking accuracy, top-k nodes, and distinguishing ability. Furthermore, HSI exhibits high consistency in evaluating node outbreak sizes when compared to SIR simulations. Our method offers valuable insights that can be leveraged for network management and the development of intervention strategies.