Machine Learning-Based Phishing E-mail Detection Using Persuasion Principle and NLP Techniques

At present time phishing attack is one of the biggest problem in our regular life. Phishing email is a type of cyberattack, which is a serious problem for financial damage in commercial organization. In phishing email used a special trick to the user that can access his digital assets. Generally phishing attack targeted to normal users by sending email that contain malicious links which are used to spoofed websites, by which the attacker can collect different sensitive information. This paper proposes a machine learning base comparison approach using persuasion principle and Natural Language Processing (NLP) base features (TF-IDF). Here we use 7 different classifiers for phishing email detection. We prepared a emails dataset collected from the Enron corpus and the Nazario phishing email corpus which are well-known and the most popular. We performed the comparative result analysis of different classification algorithms and shown the heighst performance with 99% accuracy value of AdaBoost classifier for phishing email detection.