Ensemble-Based Dynamic Phishing Domain Detection in Web Browsing Extensions

Abstract

In the continually changing world of cybersecurity threats, this study aims to improve the effectiveness of dynamic phishing domain identification within online surfing extensions. Phishing attacks use deceptive tactics to exploit vulnerabilities introduced by popular browser extensions. Recognizing the importance of adaptive and robust defences, our study conducts a thorough review of several ensemble strategies to determine the most effective strategy. Ensemble approaches, well-known for their ability to combine several models, pro- vide a strategic response to the dynamic and ever-changing nature of phishing attempts. The primary goal of this study is to examine and compare several ensemble approaches, such as bagging, boosting, random forest, stacking, ensemble of ensembles, and gradient boosting. Accuracy, flexibility, computing efficiency, and interpretability are all carefully assessed to provide a complete picture of each technique’s strengths and weaknesses. Our findings not only shed light on the complex intricacies of ensemble approaches, but also provide a practical guidance for selecting and implementing appropriate models for dynamic phishing domain detection. The study emphasizes the need of adaptive cybersecurity solutions in combating the constant evolution of cyber threats. We hope to harden web surfing extensions against phishing attackers’ complex strategies, resulting in a more secure online experience. In conclusion, this study advances cybersecurity practices by providing actionable insights into ensemble-based techniques for dynamic phishing domain identification. As the digital landscape continues to present new difficulties, our work strives to provide cybersecurity practitioners with effective tools and techniques for building a resilient defence against the ever-changing threat landscape.