Improving information centrality of a node in complex networks by adding edges
Abstract
The problem of increasing the centrality of a network node arises in many practical applications. In this paper, we study the optimization problem of maximizing the information centrality of a given node in a network with nodes and edges, by creating new edges incident to . Since is the reciprocal of the sum of resistance distance between and all nodes, we alternatively consider the problem of minimizing by adding new edges linked to . We show that the objective function is monotone and supermodular. We provide a simple greedy algorithm with an approximation factor and running time. To speed up the computation, we also present an algorithm to compute -approximate resistance distance after iteratively adding edges, the running time of which is for any , where the notation suppresses the factors. We experimentally demonstrate the effectiveness and efficiency of our proposed algorithms.
Cite
@article{arxiv.1804.06540,
title = {Improving information centrality of a node in complex networks by adding edges},
author = {Liren Shan and Yuhao Yi and Zhongzhi Zhang},
journal= {arXiv preprint arXiv:1804.06540},
year = {2018}
}
Comments
7 pages, 2 figures, ijcai-2018