| 基于本体论的E-MAIL分类方法 |
| |
| 引用本文： | 林夕伟.基于本体论的E-MAIL分类方法[J].浙江教育学院学报,2005(2):63-68. |
| |
| 作者姓名： | 林夕伟 |
| |
| 作者单位： | 浙江教育学院,信息学院,浙江,杭州,310012 |
| |
| 摘 要： | 为了方便电子邮件的管理,用户会将邮件分入有一定涵义的类别之中,现有分类方法精度不高.因此提出一种基于本体论的E-MAIL分类方法,在本体的辅助下,对E-MAIL文本进行分析,使用"泛化扩展"方法引入相应的背景知识.然后运用RIPPER方法进行规则提取.本分类法使分类更为精确,减少了所需训练文本的数量.
|
| 关 键 词： | 本体 电子邮件分类 机器学习 背景知识 |
| 文章编号： | 1671-6574(2005)02-0063-06 |
| 修稿时间： | 2004年11月9日 |
An Ontology-based Method for E-MAIL Classification |
| |
| LIN Xi-wei.An Ontology-based Method for E-MAIL Classification[J].Journal of ZHEJIANG Education Institute,2005(2):63-68. |
| |
| Authors: | LIN Xi-wei |
| |
| Abstract: | Most E-MAIL users would classify E-MAILs into some meaningful categories, and the classification rules are closely related to a user's knowledge. Many old E-MAIL classification methods are hard to improve the performance of E-MAIL classification.An ontology based E-MAIL classification method is presented in the paper. Using ontology, users' background knowledge can be integrated with E-MAIL text, and the RIPPER method can be applied to inducting rules. Ontology-based method makes E-MAIL classification more accurate, and reduces the amount of training involved. |
| |
| Keywords: | |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
