Mail Spam Detection Using Stacking Classification

Spam mails are very fast growing and costly problem, that becomes a big trouble now-a days as they are very dangerous to recipients. They cause a lot of problems such as waste of storage space, reduction of communication band width and time losing for the identification and removal of their causes. In this paper a machine learning technique of two proposed stacked configuration will be applied on email data set. This data set has two types of emails, ham mails and spam mails. The preprocessing of these mails based on the analysis of all parts that constitute an email. Rather than considering only one part of an email such as content (mail body). The results of the proposed algorithm will be analyzed based on the training and testing of various performance evaluation metrics. Finally a comparative study will be applied with some of the recent models developed for this subject.