The Research of E-mail Classifying System Based on Rough Set
Cheng Ke-hui · Journal of Anhui University of Science and Technology · 2007
E-mail does much good to people,but at the same time,too many spam e-mails also bring us destructions.This article discussed a new way of compressing the characteristic dimension of the mail collection.Based on the reduction method of the best attribute of the dependent rough set and non-affection of the classification ability of the condition attribute,it uses the reduction-decision form of the rough set,which then results in a rule.Using the rule,it will classify the newly-joined mails.The result of experiment shows that this method is effective.