Research in a Method and Model of Spam Filtering based on Bayesian Classifier
Lizhong Tu · Journal of Nanjing Normal University · 2006
The increasing junk mail brings great inconvenience and danger to people, threatens the safety of the network. The filtering way is single used by traditional filters, can't well satisfy the demand of filtering. This paper has analysed the key techniques and methods about Bayesian classifier of content-based, provided the effective way and process of kernelly arithmetic in filtering and completed the judgment of spam. In order to reducing the damages because of mistaking e-mail, we provide the improved methods of using the risk minimization Bayesian decision and self-improvement of categorization system. The paper finally has described a spam filtering model and process by double defending based on rule and content.