A Novel Phishing Attack Detection System Based on Web Platforms by Using Deep Learning Methodology

Global economic expansion facilitated by technology has led to an evolution in phishing attacks in recent years. An uptick in phishing and other sophisticated online attacks, together with an uptick in impersonation frauds, has contributed to a general uptick in fraudulent activity losses this year. To safeguard users' online activities, a more effective phishing detection mechanism is necessary, since the worldwide impact of these attacks will only grow. Using the URL and website content such photos, text, and frames, this study aimed to build a deep learning-based phishing detection system to fill this demand. When it comes to image and natural language processing, deep learning methods work wonders. Binary Learning with Phishing Detection Strategy (BLPDS) is a novel approach to efficiently detect phishing attacks introduced in this paper. To assess its efficacy, it is cross-validated with the conventional Long Short-Term Memory (LSTM) method. Applying the BLPDS to big datasets allowed for a thorough experimental evaluation and comparison of its efficacy in detecting phishing web pages and attacks. The precise proof demonstrating the efficacy of the suggested system is shown in the result section.