MACHINE LEARNING BASED WEB PHISHING CLASSIFIERS FOREVASION ATTACKS AND DEFENSE MACHANISMS

Phishing is an electronic fraud through which an attacker can access user credentials.Phishing websites are the ones that mimic legitimate websites.Fraudsters can replace them within hours to evade their detection.The effects of phishing attacks exhibit the need for antiphishing mechanisms.There were various methods for identifying phishing websites, including heuristic-based, machine learning, blacklist, and white list approaches.Although this issue has only been investigated on a small scale, prior research has demonstrated that classifiers may be vulnerable to evasion attempts.As a result, the study discusses evasion assaults and how they are detected in the rarely examined setting of website classifiers.The suggested method addresses these shortcomings by classifying webpages using a variety of machine learning techniques and extracting information from URLs.The process entails creating adversarial samples that target categorization features while preserving the look and feel of phishing websites.The appearance is evaluated using image distortion metrics named mean squared error.Then a resemblance approach is utilized for the aim of detecting assaults that happened as a result of evasion attacks.This research introduces a novel defense mechanism against evasion attacks, marking a significant contribution to the field.