Abstract:

Websites in today's world serve several functions.Web security worries are on the rise along with the daily increase in internet users. Cyber attacks on individuals are becoming more and more commonplace. Phishing is among the often occurring web attacks. Phishing is a social engineering attack method that is frequently employed to acquire user-sensitive data, such as login credentials, credit and debit card information, and so forth. Phishing websites mimic the name and design of a legitimate website. Commonly known as a fake website, it tries to trick visitors into giving up their identities. Maximizing user protection against phishing websites was one of the main objectives in developing these models. With clever phishing detection management techniques, designers can contribute to the achievement of this objective. In this study, we describe an unique method for detecting phishing websites on the client-side using a machine learning algorithm. We use the extraction framework rule in this system paper to extract a website's attributes from just its URL. The proposed method makes use of a dataset containing 30 different URL attributes, which the same Multilayer Perceptron Classification machine learning model would make use of to evaluate the legitimacy of the website. 11,055 tuples make up the dataset used to train the model. The proposed approach results in a strong performance on the 80:20 split ratio.

Description:

Phishing, Cyber Security, Machine Learning, Multi-Layer Perceptron(MLP), Fraud Detection, Neural Network, Sensitivity Analysis