Intelligent model for classification of SPAM and HAM

Abstract

In our study, we propose a collaborative approach by using cluster computing with the help of parallel machines for fast isolation of SPAM and HAM. A cluster approach can increase the computing power many folds with existing hardware and resources thus by increasing the speed of processing without incurring any extra cost. In this study, we only use header based filtering method, thus by keeping the privacy of the user intact. The standard test set for HAM and SPAM from Spam Assassin [1][2] is used. Two types of parallel environments are used in this research. First is where multiple Anti Spam methods are used in the parallel environment against the test corpora and false positive and false negative accuracy recorded. The second parallel environment is where standard test corpora are divided into parts and fed into parallel machine environment with single anti spam method used at all machines and the time saving is recorded against standalone machine being used. Weka Data Mining Software is used to apply the anti-spam methods (available at http://www.cs.waikato.ac.nz/~ml/weka/) [3].