Can We CAN the Email Spam

Abstract

The purpose of email spam is to advertise to sell, phishing attacks, DDOS attacks and many more. Many solutions of various kinds such as blacklisting, whitelisting, grey-listing, content filtering have been proposed at the sender and receiver levels. There has been some level of success but the email spam still hits the inbox and more so the problem is false positives and false negatives. The current filtering solutions used are mostly a combination of few of the available techniques, the most common being a combination of listing and content filtering techniques. Apart from any attacks, email spam causes a great loss in resources and productivity of the user. Apart from this, this problem of false negatives exists as there is non-existence of filters to address user's preferences and behaviors, timing of the day and year. The filters at the mail servers are trained with spam and ham training data that is generic in nature. Hence, there is need to address this problem. This paper aims to address all of the above i.e. an email that hits a particular users inbox by escaping the mail server spam filtering solutions. To do this, this paper describes the problem of spam, followed by the filtering mechanisms, techniques, learning email filter model and then proposes a model to fine tune the filter to increase the efficiency.