Improving Spam Email Filtering Systems Using Data Mining Techniques

Improving Spam Email Filtering Systems Using Data Mining Techniques

Wasan Shaker Awad, Wafa M. Rafiq
ISBN13: 9781799824183|ISBN10: 1799824187|EISBN13: 9781799824206
DOI: 10.4018/978-1-7998-2418-3.ch003
Cite Chapter Cite Chapter

MLA

Awad, Wasan Shaker, and Wafa M. Rafiq. "Improving Spam Email Filtering Systems Using Data Mining Techniques." Implementing Computational Intelligence Techniques for Security Systems Design, edited by Yousif Abdullatif Albastaki and Wasan Awad, IGI Global, 2020, pp. 43-72. https://doi.org/10.4018/978-1-7998-2418-3.ch003

APA

Awad, W. S. & Rafiq, W. M. (2020). Improving Spam Email Filtering Systems Using Data Mining Techniques. In Y. Albastaki & W. Awad (Eds.), Implementing Computational Intelligence Techniques for Security Systems Design (pp. 43-72). IGI Global. https://doi.org/10.4018/978-1-7998-2418-3.ch003

Chicago

Awad, Wasan Shaker, and Wafa M. Rafiq. "Improving Spam Email Filtering Systems Using Data Mining Techniques." In Implementing Computational Intelligence Techniques for Security Systems Design, edited by Yousif Abdullatif Albastaki and Wasan Awad, 43-72. Hershey, PA: IGI Global, 2020. https://doi.org/10.4018/978-1-7998-2418-3.ch003

Export Reference

Mendeley
Favorite

Abstract

Email is the most popular choice of communication due to its low-cost and easy accessibility, which makes email spam a major issue. Emails can be incorrectly marked by a spam filter and legitimate emails can get lost in the spam folder or the spam emails can deluge the users' inboxes. Therefore, various methods based on statistics and machine learning have been developed to classify emails accurately. In this chapter, the existing spam filtering methods were studied comprehensively, and a spam email classifier based on the genetic algorithm was proposed. The proposed algorithm was successful in achieving high accuracy by reducing the rate of false positives, but at the same time, it also maintained an acceptable rate of false negatives. The proposed algorithm was tested on 2000 emails from the two popular spam datasets, Enron and LingSpam, and the accuracy was found to be nearly 90%. The results showed that the genetic algorithm is an effective method for spam classification and with further enhancements that will provide a more robust spam filter.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.