Technical Research On Behavior Pattern Recognition Spam Under The Campus Net Environment

Email is becoming one of the necessary ways for the communication of people in their daily lives because of its convenient and the low cost characteristics. But following question of the spam is growing day by day, especially in campus. How to prevent the proliferation of spam effectively is core question in the network construction which must be pay attention to. At the same time, it is also an important question in the campus net developing process which waits to be solved urgently.This paper gives a brief introduction at first of domestic and foreign present situations on spam, its harm, and category and so on. Pointed out its proliferation in the campus, has affected young people's physical and moral integrity seriously. Then described the email transmission process, the principle, the correlation protocol in detail, and pointed out the SMTP protocol loophole is the main fountainhead of spam; The following closely introduced at present domestic and foreign anti-spam filtration technology. Had pointed out the mainstream status was based on the content filtration technology at present, but along with technology process going ahead, it displays more and more flaws. The behavior pattern recognition technology arises at the historic moment. It becomes spam filtration engineering research's focus day by day. It is aims at spammer to demonstrate in the transmission process the behavior characteristic carries on received, processed, judged simultaneously. It received, processed and judged the spam simultaneously according to the behavior characteristics in the process of transmission. This technology greatly enhanced the gateway filtration spam speed, reduced the network resource's load.From massive primitive mail diary, I used the data mining technology, discovered more behavior characteristics of spam transmission, and used the C4.5 algorithm to find the determination rule from the decision tree. Finally I have carried on test, and compared result with the result of school original mail filtration system. Draw a conclusion: My research may obtain more accurate interception on spam. All primitive mail diary which uses in the paper is authenticity. Spam filtration technology development direction will certainly to be the behavior pattern recognition technology primarily, other kinds of technologies be auxiliary. And they work together to solve the spam problem.