改进动态克隆选择算法在邮件过滤中的应用研究  

Research on an Improved Dynamic Clonal Selection Algorithm Applied to E-mail Filtering

作　　者：麻小强[1] 伍忠东[1] 王静[1] 王轶群[1] 

机构地区：[1]兰州交通大学电子与信息工程学院,甘肃兰州730070

出　　处：《兰州交通大学学报》2007年第6期103-106,共4页Journal of Lanzhou Jiaotong University

摘　　要：提出了可控变异和随机变异方法,改进了动态克隆选择算法,建立并实现一种改进的基于人工免疫系统的邮件过滤算法.并采用SpamAssassin垃圾邮件样本对算法性能进行测试.实验结果表明,引入改进的动态克隆选择算法后,该算法对具有变异特性的垃圾邮件能保持较高的正确识别率,具有较好的自适应性和多样性.An improving algorithm of e-mail filtering based on the artificial immune system is established and realized, the methods of the controllable-aberrance and random-aberrance are put forwar, and the dynamic clonal selection algorithm is improved. Furthermore, a spam sample called SpamAssassin is applied to testing the performance of algorithm. The result of the experiment shows that the algorithm is able to maintain a higher correct recognition rate for spam which is the character of aberrance after introducing the improved dynamic clonal selection algorithm,and it has better adaptability and multiplicity.

关 键 词：动态克隆选择算法 垃圾邮件 人工免疫 抗原/抗体 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]