Detection of Phishing Websites Using Machine Learning

Phishing attacks costs internet users billions of dollars across the globe. Phishers usually target naive users by sending them phishing links via email or other messaging services. Phishers steal financial account details, personal information, account username and password from their victims. Traditionally there are different methods to detect phishing websites like visual confirmation, content-based approach, using blacklist of phishing URLs, etc. But it is getting difficult to detect phishing websites due to reasons like rise in phishing websites every day, using URL obfuscation to shorten URL, redirecting links and manipulating links to make them look trustable. In this paper we propose a machine learning based approach that will take into consideration all these complications and detect phishing site. The goal of this paper is that, we will find out which machine learning model and its parameters are best suited to detect phishing sites at high accuracy.