支持向量机的中文电子邮件作者识别研究  

Analysis on Author Identification in Chinese E-mail by SVM

作　　者：蒲治宇 高源 Pu Zhiyu;Gao Yuan(School of Computing,Sichuan Technology and Business University,Chengdu 611745 China;Key Laboratory of Cloud Computing and Intelligent Information Processing,Sichuan Technology and Business University,Chengdu 611745 China)

机构地区：[1]四川工商学院计算机学院,四川成都611745 [2]四川工商学院云计算与智能信息处理重点实验室,四川成都611745

出　　处：《四川工商学院学术新视野》2018年第4期71-74,88,共5页Academic New Vision of Sichuan Technology and Business University

摘　　要：电子邮件已经成为当今人们最重要的通信方式之一。但随之而来的电子邮件骚扰、诈骗等现象日益严重,对邮件作者的识别有助于营造绿色、安全的电子邮件通信环境。文章首先提出向量空间模型针对电子邮件作者的写作风格提取特征向量,然后运用支持向量机算法构造作者风格分类器,从而构造出作者身份识别模型。最后,实验结果显示识别准确率达到95%以上,该方法在识别中文电子邮件作者身份上具有较高的可靠性。E-mail has become one of the most important means of communication.However,the following e-mail harassment,fraud and other phenomena are becoming more and more serious.Identifying the author of the e-mail is helpful to create a green and secure e-mail communication environment.In this paper,we first propose a vector space model to extract feature vectors for e-mail authors'writing styles,and then construct an author's style classifier by using support vector machine(SVM)algorithm,so as to establish an author's identity recognition model.Finally,the experimental results show that the recognition accuracy rate is more than95%.This method possesses high reliability in identifying Chinese e-mail authors.

关 键 词：支持向量机 身份识别 电子邮件 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]