Study Of Discriminate Model For Spam Character

E-mail is the most economical and faster communication method, while spam has become a difficult problem over internet Anti-spam technology has been developing for tens of years. There are filtering models of client filtering and server filtering. As for ways of dealing with mails, there are manual filtering and automatic filtering. The automatic filtering includes automatic filtering based rules and automatic filtering based probability. Automatic filtering based probability is mainstream. But manual filtering is also used in mail filtering system as the first checking point for the reason of speed and bandwidth. Automatic filtering technologies based rules and probability is gradually practically used in e-mail systems. Bayesian algorithm, vector support algorithm and artificial intelligence algorithm are appearing. These technologies have their advantages while there is not a valid one. Lack of a relatively complete technical definition of spam created the dilemma which is far from the expected target.Many professional anti-spam softwares are recommended by those companies, and multi-filtering algorithm is mainstream.Generally saying , they synthesize all the above anti-spam technologies and are divided 'basic filtering layer' and 'high-grade filtering layer'. The former includes black/white list and black/white keywords, the latter includes four layers of anti-spam safe-box, study engine based contents, rule engine based behavior identification, blacklist accepting engine. The multi-defense system, which considers speed and seeks for the best effect of filtering, is formed.The paper will sum up the technology characters of all sides and draw a more satisfactory description of spam with using the knowledge of statistics, vector space model and vector analysis. We will construct a spam discriminate model on the base of the character of email protocol and the spam definition of Chinese Internet Association. We will also define the technological characteristic of the spam, and classify thetechnological characteristic of spam, in order to stress its key level relation and define every characteristic importance and interrelation when spam is judged. The basis for spam of technical treatment is offered to reduce the phenomenon, which attending to one thing and lose sight of another, appearing frequently in the handling of spam.An effective anti-spam technology solution proposal based the model will be brought out, which is going to create theoretical foundation in order to solve the problem of the spam finally.It is creative that spam is studied, and mathematical descriptions are given, which is a discriminate model for spam.