Content Based Phishing Detection with Machine Learning

In recent years due to the inevitable growth of Internet technologies, almost all the real-world systems are transferred to digital platforms. This increases the use of cyberspace in every dimension of our lives especially with mobile devices which enable us to connect to related services in anytime and anywhere concept. However, this ineluctable expansion also brings lots of security breaches especially for standard end users. Phishing is one of the mostly preferred attack types that hackers use by easily hindering themselves. This type attack is initially triggered with a simple e-mail or social media message which mainly forward the victims to a malicious webpage. For security admins, they are really hard attack types to detect. Therefore, in this paper a content based phishing detection mechanism is proposed. In the proposal about six different machine learning models are implemented to select the best training models. Experimental results show that the proposed approaches are very robust and give acceptable accuracies for security admins.