Chapter

Networked Digital Technologies

Volume 293 of the series Communications in Computer and Information Science pp 78-89

Combining Classifiers for Spam Detection

  • Fatiha BarigouAffiliated withComputer Science Laboratory of Oran Computer Science Department, Faculty of Science, University of Oran
  • , Naouel BarigouAffiliated withComputer Science Laboratory of Oran Computer Science Department, Faculty of Science, University of Oran
  • , Baghdad AtmaniAffiliated withComputer Science Laboratory of Oran Computer Science Department, Faculty of Science, University of Oran

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Nowadays e-mail has become a fast and economical way to exchange information. However, unsolicited or junk e-mail also known as spam quickly became a major problem on the Internet and keeping users away from them becomes one of the most important research area. Indeed, spam filtering is used to prevent access to undesirable e-mails. In this paper we propose a spam detection system called “3CA&1NB” which uses machine learning to detect spam. “3CA&1NB” has the characteristic of combining three cellular automata and one naïve Bayes algorithm. We discuss how the combination learning based methods can improve detection performances. Our preliminary results show that it can detect spam effectively.

Keywords

spam cellular automaton Naïve Bayes classifier combination