Spam Email Detection using Naïve Bayes classifier

Abstract

Spam email detection is still a considerable and ongoing challenge in today’s online environment, as the number of unsolicited emails keeps growing exponentially. Various algorithms such as the tree-based model, support vector machine Algorithm, and Convolutional Neural Network have been explored in prior research to tackle this challenge. This research specifically examines the effectiveness of the Naïve Bayes classifier for identifying and filtering spam emails. By delving into the fundamental principles of this classifier, its practical implementation, and the comprehensive evaluation of its performance on a combined dataset, its strengths and limitations in distinguishing spam from ham messages are revealed. The result of the study demonstrates an overall accuracy of 97.82%, showcasing the Naïve Bayes classifier's high efficiency and stability in identifying spam. With consistently high metrics score throughout both classes, the Naïve Bayes classifier has proven to be an exceptionally reliable tool for spam email detection, underscoring its suitability for numerous real-world applications.