An Efficient Detection of Phishing Website using Machine Learning

Phishing is the practice of creating fake websites to track and steal sensitive information of online users. It is a type of identity theft where thieves create fake accounts on targeted websites and tricking victims into revealing their personal information like passwords, PINs and more. Phishing attackers send phishing links via email, text messages and emails, social media. They use social engineering techniques to lure customers to phishing websites and provide sensitive personal information. Finally, stolen personal information is used to establish trust between legitimate websites or financial institutions for illegitimate gains. Enormous amount of information is simultaneously downloaded and published on the web. This makes it possible for criminals to seize sensitive personal information. Use of machine learning to solve these problems by developing an intelligent, efficient and flexible system. Phishing websites can be detected based on important characteristics like URL and registration number. Many ecommerce businesses can use this system to simplify the business process. Thanks to this system, in which users can purchase online without any hassle. Information is presented and reviewed. A website can be classified as legitimate or phishing using supervised classification algorithms such as logistic regression, gradient boosting, decision trees, and support vector machines. The proposed model achieved $97.4 \%$ accuracy in gradient enhancing classifiers among the other algorithms compared.