基于DTRS模型的邮件过滤方法研究

ON DTRS MODEL-BASED EMAIL FILTERING

邮件过滤是当前网络信息安全研究的一个热点。针对传统邮件过滤方法容错能力方面的不足,提出一种基于决策粗糙集模型DTRS(Decision-Theoretic Rough Set)的邮件过滤方法。通过将无法明确判断的邮件用DTRS的边界域进行刻画,实现正常邮件、垃圾邮件和可疑邮件的三枝决策,确保总体决策的完备性。仿真实验结果表明文中方法是有效的,并且在控制邮件误分类上具有优势。

Email filtering is a research focus in regard to current networks information security.Aiming at the weak fault-tolerant ability of traditional email filtering methods,we propose an email filtering approach which is based on DTRS（decision-theoretic rough set）.A three-way decision for normal,spam and suspicious emails is realised by depicting those emails incapable of explicit judging with boundary region of DTRS,and this ensures the completeness of overall decision.The results of simulation experiment show that this approach is effective and has the advantage in controlling the email misclassification.