Paper Key : IRJ************230
Author: Lingeswar Kb
Date Published: 17 Apr 2024
Abstract
Phishing attacks pose a significant and increasing risk to internet users, resulting in substantial financial losses annually. To address this threat, this study presents a novel machine learning model aimed at detecting deceptive phishing URLs. Drawing upon data from PhishTank and a university dataset encompassing both legitimate and malicious URLs, the researchers developed and trained the model using over 5,000 carefully selected URLs. These URLs were meticulously split for training and testing, ensuring a balanced representation of both legitimate and phishing links. The model examines various features within URLs, including characteristics from the address bar, domain, and HTML and JavaScript code. By identifying discernible patterns present in phishing URLs, the model can effectively differentiate them from legitimate ones. This innovative approach is tailored for seamless integration into web applications, enabling real-time analysis of URLs to identify potential phishing attempts and safeguard users from falling prey to online scams.
DOI LINK : 10.56726/IRJMETS52416 https://www.doi.org/10.56726/IRJMETS52416