ARTIS人工免疫模型在邮件过滤中的研究与应用

Research and Application of Spam Filtering Based on the ARTIS

将人工免疫思想引入邮件过滤中,设计并实现了一种基于ARTIS(Artificial Immune System)人工免疫模型的反垃圾邮件模型.该模型将垃圾邮件看作侵入系统的抗原,模拟抗体消灭抗原的机理,以分布式方式识别垃圾邮件,并能学习和记忆邮件的特征.利用CCERT的邮件样本集对该模型进行了训练和测试,实验结果表明该系统具有较好的自适应性和稳定性.

The paper adopts the idea of Artificial Immune to classify E-mail, and build a Anti-Spam model based on ARTIS（Artificial Immune System）. In this model, every E-mail is considered as an antigen invading the system, and be detected by distributed means. The model can still learn new E-mail features. The result of experiment shows that the system possesses better adaptability and stability.