• Version
• Download 17
• File Size 186.22 KB
• File Count 1
• Create Date 24 January 2025
• Last Updated 24 January 2025

URL BASED PHISHING DETECTION

Ms.MADHU T , Ms. MONICA M Ms. SHYMA S (MENTOR)

ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING

SRI SHAKTHI INSTITUTE OF ENGINEERING AND TECHNOLOGY , COIMBATORE.

 

 

Abstract - Phishing attacks, which deceive users into revealing sensitive information by mimicking legitimate websites, pose a growing threat in the digital age. To address this challenge, we propose a machine learning- based system for detecting phishing URLs. The system uses logistic regression in conjunction with TF-IDF (Term Frequency-Inverse Document Frequency) vectorization to analyze and classify URLs as either legitimate or phishing. By identifying suspicious patterns in URL structures, such as unusual domain names, special characters, or deceptive keywords, the model effectively predicts whether a given URL is malicious. The detection system is deployed through a web interface built with Flask, allowing users to input URLs for real-time analysis. If a URL is flagged as phishing, the system provides an alert along with specific insights into the factors that led to the decision. Additionally, the system checks the URL against a database of known phishing websites, ensuring efficient recognition of already verified threats. This project provides an easy- to-use, scalable solution for combating phishing attacks by combining machine learning techniques with web integration. It aims to enhance online security for both individual users and organizations, offering a reliable tool to prevent phishing and protect sensitive information.

Download