General Automatic Learning Mail Filtering System
Renbo Pang, Li Zhang, Wei Zhang · 2008
There is a lack of general mail filtering system, which is not only compatible with several content-based filtering methods but also can realize automatic learning function by mail client. In this connection, this paper proposes GALMFS (general automatic learning mail filtering system). GALMFS realizes general design of mail filtering system through separating content-based filtering methods from mail filtering system. And it achieves automatic learning function of mail filtering system. Furthermore, it uses thread pool technology and windows overlapped I/O events communication pattern to keep efficient and stable communication between mail client and mail server. Experiments show that GALMFS can improve accuracy in spam filtering constantly by automatic learning function and meet filtering requirement of users.