The Research And Its Application Of Anti-spam Technology

Internet has become one of the most popular communication and working way in modern world. E-mails, thanks to its proficiency, convenience and low cost, have been getting more and more wildly used, and thus dominant in communication market. At the same time, however, as its unavoidable by-product, spams are bringing endless troubles to Internet users, network administrators and Internet service providers. Because of them, users' time is wasted, bandwidth is occupied, storage space is consumed, and even the Internet is blocked. Spams have been considered as the carrier to information, which has been spread everywhere. Nowadays, the mature filtering methods against spam combine the automatic software filtering and manual management. However, those methods have been proved to be unpractical to various spams, and it is estimated that only 80% of spams can be detected and filtered. Focused on those different email sending methods or modes, this thesis puts forward the simultaneous filtering strategy performed by mail server and client, which is based on the different filtering technique, and finally we are able to achieve the purpose. Used easily and filtering the most junk mail, real blackhole list technique has proved to be the most prosperous way. Installed RBL, the excessive loading of server can be decreased greatly. The rules matching technique can actualize the rule against spam. This thesis, for instance, has designed hundreds of such rules, and each one has a weight. If a mail outweighs its estimated one, it will be detected as a spam. The Bayesian statistics is the basic method of the statistics filtering. It can produce the personal filtering scheme according to different mail clients after efficient learning. The strongpoint of the personal filter is different to each other. If each filter has different statistics weight, optimized system used by the spam senders will run in an amazing slow way. Though varieties of filtering methods exist now, a large number of problems about spam filtering still remain to be solved, which impedes the filtering performance. Especially, the spam senders change their ways of sending, which prevents the filtering from functioning. Therefore, it is necessary to spend more time studying the spam filtering.