Advancements in Phishing Detection: A Comprehensive Review of Deep Learning and Hybrid Approaches

Phishing attacks continue to be one of the most prevalent forms of cybercrime, leveraging social engineering techniques to trick users into disclosing sensitive information or unknowingly installing malicious software. These attacks often rely on deceptive URLs that lead victims to fraudulent websites designed to steal login credentials, financial data, or personal information. Despite numerous studies focused on detecting phishing through HTML and URL-based methods, there is a lack of comprehensive surveys that systematically review these approaches. This paper addresses this gap by offering an in- depth analysis of modern deep learning techniques applied to phishing detection, with a particular emphasis on URL-based and hybrid methods. The study examines key aspects such as data preprocessing, feature extraction, model architecture, and overall performance, while also highlighting research gaps in the application of deep learning to phishing detection. This paper provides valuable insights for both researchers and cybersecurity professionals, offering a roadmap for future developments in phishing detection technologies.