全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于Boosting算法的垃圾邮件过滤方法研究

Keywords: Boosting算法,垃圾邮件,过滤,分类器,评价

Full-Text   Cite this paper   Add to My Lib

Abstract:

为解决垃圾邮件过滤的精确度和有效性问题,提出了一种基于邮件内容过滤的垃圾邮件过滤方法,该方法采用Boosting算法构造了一种垃圾邮件过滤器,利用该垃圾邮件过滤器实现对垃圾邮件的过滤.本文借鉴文本分类和信息检索领域所使用的评价指标,构建了垃圾邮件过滤器的评价体系,利用该评价体系,针对基于Boosting算法所构造的垃圾邮件过滤器对垃圾邮件的过滤实验所得到的数据进行了测试和评估,测试和评估的结果验证了Boosting算法在垃圾邮件过滤中的有效性,其性能优于传统的贝叶斯算法.

References

[1]  Valiant L G. A theory of leamable[J]. Communication of ACM, 1984,27:1134-1142.
[2]  Schapire R E. The strength of weak leamability[J]. Machine Learning, 1990,5:197-227.
[3]  Freund Y. Boosting a weak algorithm by majority[J]. Information and Computation, 1995,121(2):256-285.
[4]  Freund Y,Schapire R E.A decision-theoretic generaiization of online learning and an application to Boosting[J].Journal of Computer and System Sciences, 1997,55(1):119-139.
[5]  Duffy N,Helmbold D.Boosting methods for regression[J]. Maehine Learning, 2002,47(2-3):153-200.
[6]  杨峰,曹麒麟,段海星,等.基于DNS Blocklist的反垃圾邮件系统的设计与实现[J].计算机工程与应用,2003,7:11-12,45. Yang Feng,Cao Qilin,Duan Haixin, et al. Design and implementation of an anti-spam system based on DNS blocklist[J].Computer Engineering and Applications,2003,7:11-12,45. (in Chinese)
[7]  潘文峰.基于内容的垃圾邮件过滤研究 [D].北京:中国科学院计算技术研究所,2004. Pan Wenfeng.Research on content-based spam filtering [D]. Beijing: Institute of Computing Technology, Chinese Academy of Sciences, 2004. (in Chinese)
[8]  Freund Yoav, Schapire R E. A short introduction to boosting[J].Journal of Japanese Society for Artificial Intelligence, 1999,14(5):771-780.
[9]  曾春,邢春晓,周立柱.基于内容过滤的个性化搜索算法[J].软件学报,2003,14(5):999-1004. Zeng Chun,Xing Chunxiao,Zhou Lizhu. A personalized search algorithm by using content-based filtering[J]. Journal of Software,2003,14(5):999-1004. (in Chinese)
[10]  Andnoutsopoulos I, Koutisias J, Chnadrinos K V, et al. An evaluation of Narve Bayesine anti-spam filtering [C]//Proceedings of the 11th European Conference on Machine Learning. Bracelone, Spain:, 2000:9-17.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133