A Survey on the Various Machine Learning Methods of Phishing URL Detection

One of the most significant cybersecurity threats in the contemporary digital environment that are also based on the prevalence of online services including banking, ecommerce and social network, is phishing attacks. Phishing attack has become almost routine. Fake site, unknown and malicious links and people get trick into it and give their personal info. As our lives move online, the old methods about spotting scam just does not fix it anymore. That is where machine learning is being use. It goes through large amount of data and picks up the pattern and clues, and detect phishing link before you even spot anything wrong. This review provide detail of machine learning for detecting phishing site. This uses algorithm like Decision Tree, Support Vector Machine, Naive Bayes and Random Forest. Deep learning models can also be used to determine the link and websites against phishing. We review on how each method compare and which feature they work on, where they work well and in which condition they fall. The review also covers the datasets used by researcher and how they actually put these tools to the test. The findings show that machine learning and ensemble methods have the potential to represent proactive, dependable, and scale phishing detection systems and help improve the cybersecurity defense.