Improved email spam classification method using integrated particle swarm optimization and decision tree

Abstract

E-mails have become the best way to communicate formal documents over internet among users. But many people have started sending the unwanted mails to others, also called email spam. It is found that many techniques have been proposed so far to efficient mine the emails as spam or non-spammed. In existing techniques, the use of unsupervised filtering to filter the input data set is ignored by the most of the existing researchers. The use of hybridization of data mining techniques is ignored in instruct to improve the accuracy rate further for Detection of fraudulent emails. The majority of the existing techniques are limited to various significant features of emails therefore utilising more features may provide more significant results. To handle above stated limitations a new technique is proposed in this paper. The proposed technique has integrated particle swarm optimization based on Decision Tree algorithm with unsupervised filtering to enhance the accuracy rate further. The comparative analyses have clearly pointed to better results than the available techniques.