大量垃圾邮件的出现给用户收发电子邮件带来极大困扰。贝叶斯算法由于在垃圾邮件过滤处理上表现出很高的准确度,因此受到广泛应用。因此,分析常用贝叶斯算法的优缺点,并对贝叶斯算法进行改进,应用最小风险贝叶斯方法与多项式(MM)模型相结合,构成邮件分类器的基本结构,降低垃圾邮件过滤系统分类错误率,提高分类准确率。最后对反垃圾邮件技术进行总结和展望。 The emergence of a large number of spam e-mail users send great trouble. Bayesian algorithms are widely used due to their high level of accuracy in spam filtering. Therefore, the advantages and disadvantages of the commonly used Bayesian algorithm are analyzed, and the Bayesian algorithm is improved. The least-risk Bayesian method is combined with the polynomial (MM) model to form the basic structure of the mail classifier and reduce the spam filtering system Classification error rate, improve classification accuracy. Finally, the anti-spam technology is summarized and prospected.