摘要
随着电子邮件的广泛使用,垃圾邮件问题也日益严峻.基于邮件内容的过滤是当前解决垃圾邮件问题的主流技术之一.提出了一种基于带有模糊隶属度的模糊支持向量机对中文垃圾邮件过滤的方法,同时,为解决FSVM中隶属度函数的确定问题,使用了一种改进的基于类中心的隶属度函数设计方法.通过实验,使用FS-VM对垃圾邮件过滤能够取得较好的效果.
With the widespread use of e-mail,the issue of spam is also increasingly severe.Content-based filtering is one of the mainstream technologies used so far.In this paper,a method of fuzzy support vector machine with a fuzzy membership degree for Chinese spam filtering was proposed for the decision of membership function of FSVM.A new method of membership function which improves the method based on center of cluster was used.After the experiment,using FSVM,spam filter can achieve better results.
出处
《成都大学学报（自然科学版）》
2010年第2期133-136,139,共5页
Journal of Chengdu University（Natural Science Edition）
基金
重庆市教委科技计划基金资助项目(KJ090823)
