Advancements in Email Spam Detection: A Systematic Review of Machine Learning and Deep Learning Techniques
Rahul Pachare · International Journal for Research in Applied Science and Engineering Technology · 2025
Email spam continues to be a severe challenge, compromising productivity and security industry-wide. This systematic review surveys developments in machine learning (ML) and deep learning (DL) methods for detecting spam, reviewing 120 studies (2010–2023). While legacy approaches such as blocklists and rule-based filtering struggle against changing threats, ML/DL models—especially ensemble techniques (e.g., XGBoost) and neural networks (e.g., LSTM, BERT)—yield >95% accuracy. Important gaps include dependence on stale datasets (e.g., Enron) and computational inefficiencies. New trends such as explainable AI (XAI) and federated learning hold potential solutions. This review gives direction toward resilient, adaptive spam detection systems, highlighting the importance of standardized benchmarks and adversarial testing for the mitigation of contemporary spam strategies.