A phishing recognition framework leveraging web page code and textual features using recurrent neural network (RNN) algorithm

Phishing detection focusses on identifying one of the most common and emerging forms of cybercrime exploits user in a fake form. Manual detection strategies do not dynamic phishing attacks, particularly when criminals understand web page structures and textual content. To solve this problem, this research proposes a phishing recognition framework leveraging web page code and textual features using Recurrent Neural Network (RNN) technique in global phishing discovery that uses web page code and textual content. By adding structural attributes of code in HTML with meaning representations of textual content, the design effectively trains sequential dependencies and contextual relationships that differentiate phishing pages from lawful page. Experimental evaluation using benchmark phishing data shows that the proposed RNN-based framework performs better than (ML) Analysis. The investigations show the potential of (DL) models in improving web security, offering a scalable and strong solution for phishing detection software across diverse global cyber communities.