SPAM EMAIL DETECTION USING BAYESIAN BELIEF NETWORK.

Authors

  • Umoru Ibrahim Department of Computer Science, University of Abuja
  • Okechukwu Audu Edoh Department of Computer Science, University of Abuja
  • Bisallah Hashim Ibrahim Department of Computer Science, University of Abuja

Keywords:

Naïve Bayes Classifier, Bayesian network, Machine learning, Artificial intelligence

Abstract

Email is one of the most relevant means of communication in today's world. In recent time, unwanted commercial bulks emails called spam has become a large problem over the Internet. A lot of research has been carried on spam email detection using the machine learning algorithms on Enron corpus data set but there is a need of better model which can produce better prediction accuracy. This research applied Bayesian belief network (BBN) model to discover the relationship among the 13 relevant attributes of the Enron corpus spam email datasets and then classify the email into two categories (legitimate or spam). It was concluded that BBN outperformed Naive Bayes classifier, Support Vector Machine and Deep Learning Network in terms of accuracy of 99%, 98%, 95% and 85% respectively.

Downloads

Published

2021-12-31

How to Cite

Umoru Ibrahim, Okechukwu Audu Edoh, & Bisallah Hashim Ibrahim. (2021). SPAM EMAIL DETECTION USING BAYESIAN BELIEF NETWORK. International Journal of Computing, Intelligence and Security Research, 1(1), 1–16. Retrieved from http://ijcsir.fmsisndajournal.org.ng/index.php/new-ijcsir/article/view/2

Issue

Vol. 1 No. 1 (2021): International Journal of Computing, Intelligence and Security Research

Section

Articles