Adversarial Robustness of Phishing Email Detection: A Comparative Study of TF-IDF + Logistic Regression and Fine-Tuned DistilBERT
Phishing emails remain one of the most persistent cybersecurity threats, and machine-learning classifiers are widely used to detect them. Most reported detection accuracies, however, are measured on clean, in-distribution test data rather than on emails deliberately altered to evade detection. This paper reports a controlled, pairwise comparison of two phishing-detection approaches a TF-IDF + Logistic Regression baseline and a fine-tuned DistilBERT transformer trained on a unified corpus of 82,2
Researchers compared TF-IDF + Logistic Regression and fine-tuned DistilBERT for phishing email detection, finding both exceeded 98% accuracy on clean data but dropped to 64.00% and 63.64% under adversarial testing. The models showed similar vulnerability despite relying on different evidence. Adversarial testing revealed partly complementary failure modes.
Summarised by netranta from News. Open the original for the full story.
Observations (0)
Log in to add an observation.
No observations yet — add the first.