Development of an Ensemble Model for Email Filtering and Classification

• Bolaji A. OMODUNBI

  Author

• Hammed A. OLASUNKANMI

  Author

This research presents an advanced ensemble-driven framework for email classification that integrates conventional machine learning algorithms with deep transfer learning methods. A well-structured dataset comprising spam, phishing, and legitimate email samples was assembled and processed using both TF-IDF representations and contextual embeddings derived from BERT. The proposed approach combines Naïve Bayes, Support Vector Machine, and BERT models through a weighted voting strategy, with weights fine-tuned via cross-validation. Experimental evaluation based on performance metrics such as accuracy, precision, recall, and F1-score indicates notable effectiveness, with the model achieving 98% accuracy, 97% precision, 98% recall, and 98% F1-score. Furthermore, the system demonstrated a high capability in identifying phishing emails while minimizing false negative rates. These results highlight the advantage of integrating traditional machine learning techniques with deep learning models to achieve a more reliable and efficient email classification system.

• FJET_21_43_43

Copyright (c) 2026 FUDMA Journal of Engineering and Technology

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.