A Phishing URL Detection Model based on Horse Herd Optimization and Random Forest Algorithms

Phishing attacks continue to be a notable threat to network and information security. They plan to expose user information and privacy, such as login credentials, passwords, credit card numbers, and other details, by tricking internet users into thinking they are the real deal. For the detection of phishing websites, Machine learning (ML) techniques have been progressively used, because of their abilities like to learn from and adapt to complex designs and features. A new approach for detecting phishing websites using ML techniques is proposed that incorporates the URL structure. The Horse Herd Optimisation Algorithm is used to determine the features, and the suggested method is tested on a dataset of websites with phishing threats. In the context of network and information security, these techniques are employed to identify websites with phishing threats. The objectives include collecting a new dataset, extracting pertinent features, and addressing the challenges of imbalanced data and adversarial attacks in phishing detection. The findings can assist security professionals and researchers in identifying the techniques that are best suitable for improving phishing detection and prevention.