typical applications of email restriction or filtering using hot-top address, IP or domain name "black-and-white list" are: real-time blacklist (RBL) filtering combined with DNS, user-defined email white channel plus strict filtering method, etc.

2 The research on email filtering based on data mining technology uses text classification and statistical algorithms to detect spam, and Bayesian filter is the most representative one. It is self-learning, self-adaptation and extremely high accuracy, occupying the dominant position in the field of filters. Other researches include: spam detection methods based on memory information and data mining based on event feature description information.