摘要
提出一种基于决策粗糙集(decision-theoretic rough set,DTRS)模型的邮件过滤方法(spam filtering based onDTRS,SFDTRS)。首先用α-正域约简理论对邮件数据进行属性约简,然后根据最小风险Bayes决策方法,将无法明确判断的邮件用DTRS模型的边界域进行刻画,实现正常邮件、垃圾邮件和可疑邮件的三枝决策。仿真实验表明:SFDTRS方法能显著降低数据的维度,保证邮件过滤效果的同时,提高模型训练和邮件过滤的效率。
A spam filtering approach termed SFDTRS（spam filtering based on decision-theoretic rough set）is developed.Firstly,α-positive region reduction theory has been adopted to reduce mail data attributes.Then,according to the minimum risk Bayes decision theory,a three-way decision,named normal spam and doubt,is made by depicting the unjudgeable mails using boundary region of DTRS.The simulation results show that SFDTRS can significantly reduce mail data dimension,enhance mail filtering effect,and improve the efficiency of model training and mail filtering.
出处
《重庆邮电大学学报（自然科学版）》
CSCD
北大核心
2013年第1期126-131,共6页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
国家自然科学基金(61202191
71202165)~~
