基于ART神经网络的垃圾邮件过滤技术

The Technology of Filtering Spam Based on ART Neural Network

本文讨论了自适应谐振理论ART,分析了ART的工作原理,给出了ART的具体算法。以神经网络ART作为分类器来过滤垃圾邮件,ART克服了BP网络的缺点,可以对垃圾邮价进行更有效地过滤,更好的解决了垃圾邮件特征不断变化而过滤方法相对固定的矛盾。并以实例详述了ART在邮件过滤中的工作过程,获得了很好的结果。

Through the theory of adaptive resonance, we analyze ART＇ operational principle and specific algorithm. Not only filter sparns effectively,we use ART to overcome the disadvantage of BP network. , but also resolve the contradiction that the features of spams varies constantly while the filtering methods are relatively static.