Enhancing Phishing Detection: A Novel Hybrid Deep Learning Framework for Cybercrime Forensics

Abstract

Phishing attacks have emerged as a significant threat in the digital realm, exploiting both social engineering
and technological vulnerabilities to deceive users into divulging sensitive information. Traditional detection
methods often fall short in identifying sophisticated phishing attempts, necessitating advanced solutions. This
paper introduces a novel hybrid deep learning framework that synergizes Support Vector Machine (SVM),
Light Gradient Boosting Machine (LightGBM), and Multi-Layer Perceptron (MLP) algorithms to enhance
phishing detection capabilities. The system architecture comprises two primary modules: Admin and User.
The Admin module facilitates the training and deployment of machine learning models, while the User
module allows for the monitoring of blocked URLs. By integrating these components, the framework aims to
bolster cybercrime forensics and provide a robust defense against phishing attacks.