A mail discrimination study using network basis tools

Abstract

Today, many methods used for the junk e-mail (spam) discrimination. For example, blacklists and some machine learning methods. Blacklists consists of IP addresses and domain names of spammers who previously sent e-mails and caught as spam. Machine learning methods are also among many other studies in which many studies have been done in recent years. Trained machines using e-mail header information, e-mail content and sender domain information decide newly incoming e-mail messages as spam or not. In this study, a new spam filtering technique which uses machine learning with Support Vector Machine (SVM) using header information of an e-mail and some networking tools have been proposed.