Enhanced Security Measures Against Phishing Threats

The rapid advancement of digital transformation has significantly heightened cybersecurity threats on a global scale. As digitization progresses, it creates more opportunities for cybercriminals. One prevalent form of cyber threat is phishing, which aims to steal confidential user credentials. Cybercriminals use phishing tactics to gain access to an organization's digital assets and networks. Through these security breaches, they can execute ransomware attacks, gain unauthorized access, shut down systems, and demand ransoms for access restoration. Despite the existence of anti-phishing software and techniques, phishers continually find ways to bypass these defenses. Threat intelligence and behavioural analytics systems help organizations identify unusual traffic patterns, yet the most effective practice remains a comprehensive defense strategy. In this article, we introduce an approach for detecting phishing attacks through the application of machine learning (ML) methods, particularly focusing on Random Forest (RF) and Decision Tree (DT) algorithms. We utilized a widely recognized dataset of phishing attacks available on Kaggle to facilitate our ML analysis. Feature selection was performed using Principal Component Analysis (PCA) to enhance the performance of the model. The Random Forest algorithm produced results with a 92% accuracy rate. The system effectively classifies websites links, providing clear outputs indicating whether a site is classified as phishing or not.