一种基于相似度测量的新垃圾邮件发现机制

A Mechanism of New Overrun Spam Detection Based on Similarity Measure

分析新垃圾邮件发现的意义,设计用来发现新垃圾邮件的相似度测量算法——Spam-SMA,该算法使用N元字串(N-Gram)作为比较用特征,基于该算法,在规则判分的反垃圾邮件框架下,提出1种新垃圾邮件发现机制,并通过对SpamAssassin的扩展实现了该机制。在邮件服务器上进行了多次实验,结果证明,该机制可有效实现新垃圾邮件的发现。

The necessiy of a new overrun spam detection is analyzed.Then aspam detection algorithm based on similarity measure is designed, which is named Spam-SMA and it makes use of the N-gram as features for comparison.A mechanism of new overrun spam detection which uses the Spam-SMA algorithm is proposed based on the score-based rule anti-spam scheme,such as SpamAssassin,and a module of Spamassassin is implemented based on the mechanism.Experiment results show that this mechanism is effective.