The Comparison of Random Forest and XGBoost in Malicious Browser Extensions Detection in Google Chrome Based on the Feature Importance

The prevalence of malicious browser extensions poses a significant threat to users’ security and privacy in the online environment. Traditional methods of detecting such threats often fall short due to the dynamic and evolving nature of malware. In this study, we explore the efficacy of two popular machine learning algorithms, Random Forest and XGBoost, in detecting malicious browser extensions in Google Chrome based on feature importance analysis. We analyze a diverse dataset of browser extensions, extracting relevant features from their source code and behavior. The feature importance scores obtained from Random Forest and XGBoost models provide insights into the discriminative power of different features for classifying extensions as malicious or benign. Our findings contribute to the understanding of the strengths and limitations of each algorithm in detecting malicious browser extensions, offering valuable insights for improving cybersecurity measures in the web browsing ecosystem.