Phishing Detector Extension Using Machine Learning

Abstract

The goal of our project is to implement a machine learning solution to the matter of detecting phishing and malicious web links. The tip results of our project are going to be a software package which uses machine learning algorithm to detect malicious URLs. Phishing is that the technique of extracting user credentials and sensitive data from users by masquerading as a real website. In phishing, the user is supplied with a mirror website which is clone of the legitimate one but with malicious code to extract and send user credentials to phishers. Phishing attacks can cause huge financial losses for patrons of banking and financial services. the normal approach to phishing detection has been to either to use a blacklist of known phishing links or heuristically evaluate the attributes in a suspected phishing page to detect the presence of malicious codes. The heuristic function relies unproved and error to define the edge which is employed to classify malicious links from benign ones. the disadvantage to the current approach is poor accuracy and low adaptability to new phishing links. We attempt to use machine learning to beat these drawbacks by implementing some classification algorithms and comparing the performance of those algorithms on our dataset. we are going to test algorithms like Logistic Regression, SVM, Decision Trees and Neural Networks on a dataset of phishing links from UCI Machine Learning repository and pick the simplest model to develop a browser plugin, which might be published as a chrome extension.