Chinese Anti-spam Filtering System Development And Research

With the network spread widely, Email, as a communication tool, become a typical success of Internet application. Comparing with traditional mail, Email come into being just a short time, but it influence our work and life much more. However, more and more commercial, social, and political spam is a headache of its users. In fact, the number of spam is often much larger than nonspam's, so it spends much more user' s time and money.In recent years, the spam problem has already caused the domestic and international height value, all countries adopt the homologous measure to reduces the spam's influence.The technology, adopted in anti-spam system in our country at present, is mainly about rule-based filtering which utilize the rule of the spam' s header. And the content-based filtering is not mature, thus Chinese spam' s filtering effect is not very good. The the Chinese anti-spam filter system is developed and researched for the sake of filtering the spam effectively. According to the researching and analyzing to the Email theory and spam' s filtering methods, the spam' s filtering project is proposed in this paper and is implemented in this system. Rule-based filtering and risk minimization Naive Bayes are combined in this project, and this project is improved in some necessary places. The test version of this project is completed mostly under the Linux/Solaris system platform. And this software system is circulated and tested on some Internet mail server. Test result shown, this project has a better effect of anti-spam filtering.