Phishing Email Detection Model Using Deep Learning

Abstract

Email phishing is a widespread cyber threat that can result in the theft of sensitive information and financial loss. It uses malicious emails to trick recipients into providing sensitive information or transferring money, often by disguising themselves as legitimate organizations or individuals. As technology advances and attackers become more sophisticated, the problem of email phishing becomes increasingly challenging to detect and prevent. In this research paper, the use of deep learning techniques, including convolutional neural networks (CNNs), long short-term memory (LSTM) networks, recurrent neural networks (RNNs), and bidirectional encoder representations from transformers (BERT), are explored for detecting email phishing attacks. A dataset of phishing and benign emails was utilized, and a set of relevant features was extracted using natural language processing (NLP) techniques. The proposed deep learning model was trained and tested using the dataset, and it was found that it can achieve high accuracy in detecting email phishing compared to other state-of-the-art research, where the best performance was seen when using BERT and LSTM with an accuracy of 99.61%. The results demonstrate the potential of deep learning for improving email phishing detection and protecting against this pervasive threat.