Comparative between optimization feature selection by using classifiers algorithms on spam email

Abstract

Spam mail has become a rising phenomenon in a world that has recently witnessed high growth in the volume of emails. This indicates the need to develop an effective spam filter. At the present time, Classification algorithms for text mining are used for the classification of emails. This paper provides a description and evaluation of the effectiveness of three popular classifiers using optimization feature selections, such as Genetic algorithm, Harmony search, practical swarm optimization, and simulating annealing. The research focuses on a comparison of the effect of classifiers using K-nearest Neighbor (KNN), Naïve Bayesian (NB), and Support Vector Machine (SVM) on spam classifiers (without using feature selection) also enhances the reliability of feature selection by proposing optimization feature selection to reduce number of features that are not important.