摘要
针对目前大量垃圾信息充斥用户电子邮箱的问题,提出一种基于用户模型的电子邮件自动处理方法.它根据机器学习的原理,建立并且不断更新用户模型,记录用户的历史行为,并能够将历史数据作为训练样例,当训练样例达到适当的数量时,通过朴素贝叶斯分类法对新邮件进行归类,并采取相应的动作,最终帮助用户获取并保留更有价值的邮件信息.
This paper advanced a user-model based automatic method of e-mail processing. On the basis of machine learning, it creates a user model and maintains it continuously, records user＇s history action and takes the history data for training examples. When the training examples are enough, system can classify every new-arrived e-mail by naive Bayes classifier, and takes some actions according to the user model, at last, it attain an result of helping user to obtain and keep the useful information.
出处
《沈阳化工学院学报》
CAS
2005年第4期289-291,307,共4页
Journal of Shenyang Institute of Chemical Technolgy
