Spam Email Detection using Structural Features

Abstract

recent years, we have witnessed a dramatic raise in the use of web and thus email becomes an inevitable mode of communication. This is the scenario where the attackers take advantage by the mode of spam mails to the email users and misguide them to some phished sites or the users unwittingly install some malwares to their machine. This shows the importance of research activities being carried out in the field of spam mail detection. In this paper we tend to project a replacement methodology to segregate spam emails from non- spam (legitimate) emails using the distinct structural features available in them. The experiments with 8000 emails show that that our methodology preserves an accuracy of the spam detection up to 99.4% with at the most 0.6 % false positives. KeywordsDetection; Structural Feature Selection; spam classification;Machine learning application.