Phishing Website Detection using Emerging Machine Learning Models

Abstract

Due to Phishing, many users are losing their money, reputation which costs billions of dollars per year. These luring techniques are used by hackers to get personal data or information in a pond of unsuspecting internet users. Usually, hackers use the spoofed e-mail ids, various offers using phishing websites and software to get personal information of the internet users and also target confidential data like usernames and passwords, family information, contact details and emails, and many more. Every day we listen to various cyber-attacks in the world. A phishing attack is one of those cyber-attacks where the attacker develops a fake website by mimicking a trusted website. The whole aim of such websites is to steal financial assets from users. Such attacks are called phishing attacks. Phishing attack also costs the online attack and many big organizations some millions of dollars. We can eradicate this crime by taking proper countermeasures like building machine learning systems that can detect phishing more accurately. Machine learning is a tool that is showing very favorable results in every field where detection is a primary aspect. We experimented with a large dataset that has all the parameters to detect a website is a phishing one or not. We train the models for a better understanding of the detection using security IP. This parameter helps to detect phishing websites and find cybercriminals with high accuracy. In this research paper, we have used machine learning models which give high accuracy to predict whether the website is a phishing(malicious) website or if it is a safe website using the given parameters. We compared all the widely used machine learning algorithms to propose which algorithm gives the highest accuracy.