基于层间影响的时序网络节点重要性识别

Importance Identification of Nodes in Sequential Network Based on Inter-Layer Influence

现实生活中,复杂网络具有动态性和时变性,识别时序网络中的重要节点相比于静态网络更加具有挑战性。在本文中,我们提出了一种基于层间影响的超邻接矩阵在时序网络中的节点重要性识别方法。首先,我们定义了层间影响系数,其次考虑了时序网络层内和层间的关系构建了时序超邻接矩阵(TSAM)利用特征向量中心性对节点的中心性指标进行估。在多个实现数据集上的结果显示:相比经典的方法SAM以及SSAM,使用提出的算法得到的Kendall’s τ值在各时间层上均有显著提高,结果表明时序引力模型的度量对于时序网络的节点重要性度量具有十分重要的意义。

In real life, complex networks are dynamic and time-varying, and it is more challenging to identify the important nodes in the temporal network than static networks. In this paper, we propose a method to identify the importance of nodes in temporal networks based on the super-adjacency matrix based on inter-layer influence. First, we define the inter-layer influence coefficient, and then we consider the relationship between the intra-layer and inter-layer of the temporal network to construct the temporal super-adjacency matrix (TSAM) to estimate the centrality index of the node using the centrality of the feature vector. The results on multiple implementation data sets show that compared with the classical methods SAM and SSAM, Kendall’s τ obtained by using the proposed algorithm. The results show that the measurement of temporal gravity model is very important for the measurement of node importance in temporal networks.