Phishing website detection using machine learning on URL features

“Phishing is a cyber attack on internet users which makes the users to disclose their unique personal information. It targets on normal internet users to access sensitive information by means of counterfeit websites. The anonymous structure of internet helps the attackers to tamper the normal user’s information using uniform resource locator. Various phishing website detecting strategies have implemented so far. Due to a number of inefficient technologies the number of victims has grown exponentially. A strong anti phishing mechanism is required to filter out phishing attacks. In this proposed system a machine learning based website detection is introduced using URL features. A supervised machine learning method named XGBoost algorithm is employed to detect phishing URL. The proposed system is evaluated with 2000 malicious and 1000 legitimate sites, respectively. The experiments outcome shows that the performance of the proposed method is better than the recent approaches in malicious URL detection.