The Identification Of Time Balanced Influential Nodes And High Comprehensive Impact Node Groups

High-impact nodes have great influence on the structure and function of the network.Therefore,high-impact nodes identification is one of the key problems in network science.This paper will study two aspects problem of the high-impact node identificationThe first part of this paper mainly studies the time bias of the algorithm for identification of important nodes in dynamic networks.We take tow citation networks to analyze the problem.We investigated the traditional node centrality algorithms in the identification of highly influential papers,found that these algorithms have a large degree of time bias.Most algorithms tend to pick the old classic papers and ignore the new high-quality papers,because these newly papers didn't have enough time to gain the advantages,such as the number of references,compared with the old classical papers.This paper proposes a method by setting a time window for each paper and rescale the result of centricity algorithms to reduce the time bias.The nodes sorted according to their age,and for each node,we set a neighborhood radius centered on itself as its time window,so that all the nodes in the neighborhood radius have the similar age.Then calculate the score of each node with the node centrality algorithms,and rescale the scores with the time window of each nodes.Then rank the importance of each node by the rescaled score.This method can greatly reduce the time bias problem of identifying the important nodes.At the same time,compared with the traditional algorithms,this method can identify relatively new high-impact nodes in a shorter time.Two citation network data sets were used first time to analyze and verify this method,and the identification results show a strong time balance.The second part studies the recognition of node groups with high comprehensive influence.The classical methods choose influential single node,yet fail in the case of multiple ones.In the paper,we analyze the collective influence and overlapping influence of multiple spreaders.Then based on Rayleigh quotient and the overlapping influence between spreaders,we clarify the problem of why multiple influential spreaders may have low collective influence.At last,a new algorithm is proposed to choose multiple spreaders,whose performance is validated in real networks.