| 基于Logistic回归的中文垃圾邮件过滤方法 |
| |
| 引用本文： | 王庆幸,徐从富,何俊.基于Logistic回归的中文垃圾邮件过滤方法[J].计算机科学,2008,35(10):197-199. |
| |
| 作者姓名： | 王庆幸 徐从富 何俊 |
| |
| 作者单位： | 1. 浙江大学计算机学院,杭州,310027
2. 浙江省辐射环境监测站,杭州,310012 |
| |
| 基金项目： | 国家高技术研究发展计划(863计划)，国家自然科学基金 |
| |
| 摘 要： | 研究如何实现Logistic回归模型在中文垃圾邮件过滤中的应用,给出了关键技术,并将其应用于SEWM20071)垃圾邮件语料库上,取得了较优的过滤效果.还分析了影响正常邮件误判率、垃圾邮件误判率和精确率等因子.对比实验结果表明,应用于中文垃圾邮件过滤的Logistic回归模型与SVM相比具有更优的ROC指标和更快的运行效率.
|
| 关 键 词： | 垃圾邮件过滤 Logistic回归 ham％ spam％ |
Filtering Chinese Spam Email Using Logistic Regression |
| |
| WANG Qing-xing,XU Cong-fu,HE Jun.Filtering Chinese Spam Email Using Logistic Regression[J].Computer Science,2008,35(10):197-199. |
| |
| Authors: | WANG Qing-xing XU Cong-fu HE Jun |
| |
| Affiliation: | WANG Qing-xing1 XU Cong-fu1 HE Jun2(College of Computer Science,Zhejiang University,Hangzhou 310027,China)1(Radiation Environment Monitoring Station of Zhejiang Province,Hangzhou 310012,China)2 |
| |
| Abstract: | We applied the logistic regression model to filter Chinese spam email,described the key approaches of our spam filter,and conduct experiments on SEWM2007 spam corpus.Additionally,we analyzed factors influencing the ham misclassification rate(ham%),the spam misclassification rate(spam%)and the accuracy of our filter.Compared with SVM,our filter is better in terms of receiver operating characteristics(ROC)and efficiency. |
| |
| Keywords: | ROC |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机科学》浏览原始摘要信息 |
|
点击此处可从《计算机科学》下载全文 |
|
