基于统计分词的中文邮件智能分类系统  

Based on the SVM the intelligent sorting system of the Email

作　　者：徐海涛[1] 杨森[1] 柴乔林[1] 

机构地区：[1]山东大学计算机与科学技术学院

出　　处：《华中科技大学学报（自然科学版）》2003年第S1期325-328,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

摘　　要：区别于以往采取专家系统、基于语义分析及基于关键字比较的分类方法 ,根据文本数据学习的特点 ,采用支持向量机 (SVM )来实现电子邮件的智能分类方法 .通过SVM方法与其他几种分类方法试验测试 ,结果发现 ,SVM方法效果最好 。Different from the previous sorting procedures such as the expert system, text-based analysis and keyword-based comparison system, this paper adapt SVM to implement the intelligent sorting of the emails according to the text data learning characteristics. According to the tests by the SVM method and several other sorting methods, it was found that the SVM has the best performance. The experiment showed that this method can replace the human being on sorting and processing the emails by using artificial inte...

关 键 词：电子邮件 分类 支持向量机 召回率 

分 类 号：TP393.098[自动化与计算机技术—计算机应用技术]