Enhancing Email Spam Filtering through Context-aware Machine Learning Techniques

Abstract

Unsolicited email, commonly referred to as spam, remains a major security and privacy concern in digital communication. It often carries threats such as phishing attempts, malicious attachments, and unauthorized data collection. Researchers have used advanced Deep Learning (DL) techniques and traditional Machine Learning (ML) algorithms to address this problem. Naïve Bayes, logistic regression, random forest, Support Vector Machines (SVM), and Convolutional Neural Networks (CNN) are among the models that are compared in this study. Their effectiveness is measured using standard metrics accuracy, precision, recall, and F1-score. Findings reveal that while CNN achieves strong predictive performance, simpler models like Naïve Bayes remain valuable for real-time, resource-efficient spam filtering systems.