Spam filtering by stages

At the present time, the typical spam filtering technique is content filtering, but the content filtering belongs to after the event precaution, and cannot avoid network resources waste due to spam. Therefore, this paper used a method of spam behaviour recognition filtering. The method identify the spam according to the behaviour of mail sent, set up the model by Bayes technique, filter spam by the least risk of Bayes technique, and in the mail filtering application filtering spam by stages in order to recognizes the spam email as soon as possible. Experiments show that this method could get a better effect, and could identify the spam as early as possible, so that this approach could solve the problem of resources wasting.