OFMCDM/IRF: A Phishing Website Detection Model based on Optimized Fuzzy Multi-Criteria Decision-Making and Improved Random Forest

Abstract

With increasing social and financial activities on the web, phishing has become one of the most critical threats in cybersecurity. Many methods have been proposed to identify phishing websites, such as fuzzy logic, neural networks, data mining, heuristic-based phishing detection, and machine learning. On the other hand, phishers develop more sophisticated techniques, decreasing the efficacy of the existing methods. This paper proposes a phishing detection model based on optimized Fuzzy Multi-Criteria Decision-Making (OFMCDM) and Improved Random Forest (IRF). The model utilizes Uniform Resource Locator (URL) and Hypertext Markup Language (HTML) features to prevent sharing users’ sensitive information such as username, password, social security, or credit card number. Our experiments show competitive results from our models compared to existing models, including Naive Bayes (NB), Logistic Regression (LR), K-Nearest Neighbor (KNN), and Decision Tree.