Hybrid Nave Bayes Models for Scam Detection Comparative Insights From Email and Financial Fraud

The growing threat of email and financial fraud necessitates the development of efficient and scalable detection systems. This project presents a hybrid approach to fraud detection, leveraging multiple machine learning models such as XGBoost, Random Forest, LightGBM, and Naive Bayes to classify emails as fraudulent or legitimate. The project incorporates TF-IDF and BERT embeddings to extract features from the email text, enhancing the ability of the models to understand and differentiate between legitimate and phishing emails. The stacking classifier approach is employed to combine the strengths of these individual models, further improving detection accuracy. The system is evaluated using performance metrics such as accuracy, precision, recall, and F1-score. The results show that the hybrid model outperforms traditional approaches in terms of accuracy and reliability. This system provides a scalable and adaptable solution for detecting fraudulent emails, offering higher precision and lower error rates. The integration of deep learning techniques and ensemble methods positions this research as a significant contribution to the field of fraud detection.

Keywords: Hybrid Models, XGBoost, Naive Bayes, Random Forest, LightGBM, TF-IDF, BERT, Stacking Classifier, Email Fraud Detection, Financial Fraud.