Optimization of URL-based phishing websites detection through genetic algorithms

Website phishing is an online crime for obtaining secret information such as passwords, account numbers, and credit card details. Attackers lure users through attractive hyperlinks, in order to, redirect to the fake websites. Phishing detection through a machine-learning approach has become quite effective nowadays. In this research, the Uniform Resource Locator (URL) based phishing detection approach has been used. Machine-learning classifiers like Naïve Bayes, Iterative Dichotomiser-3 (ID3), K-Nearest Neighbor (KNN), Decision Tree and Random Forest used for the classification of legitimate and illegitimate websites. This classification would help in the detection of phishing websites. However, it has been observed that use of Genetic Algorithms (GAs) for feature selection can improve the detection accuracy. Our experimental results portrayed the use of Iterative Dichotomiser-3 (ID3) along with Yet Another Generating Genetic Algorithm (YAGGA) improves the detection accuracy up to 95%.