Identifying important nodes in complex networks is a key issue in network analysis. In this paper, we propose a top-k important nodes evaluation method based on Probabilistic-jumping Random Walk (PJRW). PJRW simulates the messages propagation mechanism which randomly spreads messages from different vertexes to their neighbors. In order to avoid the messages falling into local communities, a jumping probability pc is introduced. To verify the effectiveness of PJRW, Attack robustness model and Susceptible-Infected-Recovered (SIR) model are used in this paper. Experimental results on two artificial networks and seven real networks suggest that the PJRW method can effectively find the important top-k nodes and its result in most networks is better than degree, betweenness, closeness and eigenvector centrality. In addition, PJRW only uses local information and part of statistic global information of a network, its computation complexity is less than global metrics such as betweeness and closeness. It is suitable to be applied in large-scale complex networks for identifying important nodes.
CITATION STYLE
Yu, H., Chen, L., Cao, X., Liu, Z., & Li, Y. (2018). Identifying top-K important nodes based on probabilistic-jumping random walk in complex networks. In Studies in Computational Intelligence (Vol. 689, pp. 326–338). Springer Verlag. https://doi.org/10.1007/978-3-319-72150-7_27
Mendeley helps you to discover research relevant for your work.