一种新的邮件过滤技术研究

为了改进已有邮件过滤算法的不足之处，提出一种新的邮件过滤算法。以往的大部分过滤算法采用的都是邮件属性精确匹配算法，并未使用模糊匹配思想，因此邮件的漏检率较高，并且发现未知邮件的效率较低。针对以往邮件过滤算法中漏检率比较高的不足之处，文中提出的邮件过滤算法的改进思路是：首先使用传统的黑白名单过滤技术对已知的邮件进行分类，那些是正常邮件，那些是垃圾邮件。在此基础之上使用相似性算法来计算未知邮件与已知邮件的相似度，从而达到对未知邮件分类目的,最后使用MMTD算法对的邮件相似度的好坏做出衡量，并且以此为邮件过滤提供有效的依据，经过以上的三个步骤之后，最后进行邮件的过滤。

Research on a New Mail Filtering Technology

The widely used mail attributes exact matching algorithm does not use fuzzy matching ideas, so the rate of missed messages is high and the efficiency of finding the unknown message is low. This paper presented the improved spam filtering idea. First, the traditional black and white list of known e-mail filtering technology was used to classify the messages into normal e-mail and spam. Based on this, similarity algorithm was used to calculate the unknown e-mail message with the known similarity, so as to achieve the unknown message classification purposes. Then, MMTD mail similarity algorithm was used to measure the quality of e-mail message, which provides the effective basis for message filter. After the above three steps, the e-mail messages were filtered.