Overlapping community and entropy of neighborhood information for identifying influential nodes in complex networks.

Identifying influential nodes in complex networks is a fundamental problem in network science, with applications in social networks, information diffusion, and epidemic control. Traditional approaches often rely on node degree or centrality indices, which may overlook structural heterogeneity and fail to capture both local and global aspects of influence. To address these limitations, this paper proposes a novel method, OCNEI (Overlapping Community and Neighborhood Entropy-based Influence), which integrates neighborhood information entropy with overlapping community structures for detecting influential nodes. The process begins by identifying overlapping community structures, which allow nodes to belong to multiple communities and better capture the heterogeneous relationships of real-world networks. These communities are then used to guide the computation of neighborhood information entropy, where the uncertainty and diversity of each node's local connections are measured in the context of its community memberships. This entropy-based evaluation is further enhanced by considering the contributions of neighboring nodes, resulting in a more reliable metric of local influence than degree-based indices. Through this sequential integration-overlapping community detection, entropy-driven local influence, and global distance evaluation-OCNEI effectively balances precision with scalability. The effectiveness of OCNEI is validated through simulations on synthetic and real-world networks, where it achieves a 2.7% improvement under the Susceptible-Infected-Removed (SIR) model compared to the best existing method.