一种基于局部特性的重要邮箱用户发现方法

摘要

In this paper, aiming at the directed weighted e-mail networks, we put forward an important e-mail users detection algorithm based on local characteristics. The method takes into account the impact of the weight of node on node importance assessment, and closely integrating the features of the mail network, it gives new definition of the weight of node. At the same time, in order to apply to the large directed weighted network, we propose a novel algorithm of importance metric for the node based on local characteristics of the node in network, its time complexity is o(2m + n). This method has been verified on Enron E-mail corpus database, and the quite good experimental results are derived.%针对有向赋权邮件网络,提出一种基于局部特性的重要邮箱用户发现算法.该方法考虑了节点的点权对节点重要度评估的影响,紧密结合邮件网络的特性,给出了点权的新定义.同时,为了应用于大型有向赋权复杂网络,提出依据网络中节点的局部特征的节点的重要性度量算法,该算法的时间复杂度为o(2m+n).在安然邮件语料库上验证了该方法,并且得到了较好的实验结果.