Spam Filter Based on Naive Bayesian Classifier

Abstract Spam now accounts for over 70% of all emails and the harm to users is increasing, such as waste a lot of network bandwidth to transfer and space to store, has large quantity, repetitive, fraud, and unhealthy content, etc. As spam is usually embedded in normal e-mails, it is difficult to identify them. This paper analyzed the main technologies to identify and block spam, such as information or content filtering technology, blacklist and white list technology, intention and behaviour analysis technology, etc. A model to determine an e-mail is a spam or not based on naive Bayesian classifier is presented in the paper. The test result shows that the model is effective.