Phishy? Detecting Phishing Emails Using ML and NLP

Abstract

Phishing emails, a type of cyberattack using fake emails, are difficult to recognize due to sophisticated techniques employed by attackers. In this paper, we use a natural language processing (NLP) and machine learning (ML) based approach for detecting phishing emails. We compare the efficacy of six different ML algorithms for the purpose. An empirical evaluation on two public datasets demonstrates that our approach detects phishing emails with high accuracy, precision, and recall. The findings from this work are useful in devising more efficient techniques for recognizing and preventing phishing attacks.