Abstract: A method discovering important e-mail accounts based on structure is proposed in this paper, in spite of most existed methods using e-mail content as the semantic meaning to help understand. Referring to the idea of web link analysis algorithms PageRank and HITS, a method which evaluates the importance of e-mail address in e-mail graph is proposed, and also can evaluate the importance of e-mail account. The methods are verified on the Enron E-mail Dataset, and the experiment results are better than some existent methods.

Key words: EA Graph, Link Analysis Algorithm, Important E-mail Account

摘要： 多数发现邮件重要用户的方法都借助邮件的内容信息作为语义来帮助理解，本文提出了一种基于结构的重要邮件用户发现方法。该方法借鉴网页中链接分析算法的思想计算邮件地址的重要度。在安然邮件语料库上验证了我们的方法，得到较好的实验结果。

关键词: EA拓扑图, 链接分析算法, 重要邮箱用户