摘要
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.
To heighten the accuracy and effectiveness of filtering junk mail, this paper presents a mail filtering method based on filtering mail contents. The junk mail filter was constructed using Boosting algorithm. Referring to the evaluation index adopted in such fields as text classification and information retrieval, the evaluation system of presented junk mail filter was also established. The experiment and performance evaluation were implemented. Results show that, compared with traditional Bias algorithm, the performance of filtering method using Boosting algorithm is obviously superior to the method using traditional Bias algorithm.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2013年第1期79-83,共5页
Transactions of Beijing Institute of Technology
基金
黑龙江齐齐哈尔市科学技术计划项目(GYGG2010-06-02)
