Developing a Context-Aware Convolutional Neural Network (CACNN) for Enhanced Phishing Website Detection

Abstract

This paper presents the development of a Context-Aware Convolutional Neural Network (CACNN) aimed at improving the detection of phishing websites. Given the escalating sophistication of phishing attacks, traditional detection methods have become less effective, necessitating more advanced solutions. This research addresses this need by proposing a novel CACNN model that integrates visual and textual analysis of websites using advanced machine learning techniques. The CACNN model is designed to understand the context of website content, making it adept at identifying subtle cues of phishing attempts that might elude conventional detection systems. The methodology involves training the CACNN with a comprehensive dataset comprising both phishing and legitimate websites, followed by rigorous evaluation using standard performance metrics. The results demonstrate a significant improvement in phishing detection accuracy compared to existing methods, highlighting the efficacy of the context-aware approach. This research contributes to the cybersecurity field by providing a more robust and intelligent tool for safeguarding against phishing threats.