BLOCKCHAIN BASED SECURE PHISHING WEBSITE DETECTION USING MACHINE LEARNING

Abstract

Phishing websites, which prey on users' trust by seeming to be reliable online services in order to obtain private data, such as cryptographic keys, login passwords, and financial information, are among the most dangerous cybersecurity threats. The fast expansion of internet platforms, e-commerce systems, and blockchain-based apps has made phishing attempts increasingly complex, scalable, and difficult to detect. Machine learning and deep learning techniques have shown considerable potential in phishing website detection by automatically recognizing patterns from URLs, web content, and structural aspects, increasing detection automation and accuracy. However, most of the existing approaches work in centralized environments, which bring with them restrictions on trust, transparency, manipulation of results, and single points of failure. Furthermore, recent blockchain-based security research mostly focus on transaction fraud and cryptocurrency scams rather than phishing website detection with secure result verification. To solve these problems, this study suggests a machine learning-based blockchain-based safe phishing website detection system. The suggested approach ensures transparent, unchangeable, and tamper-proof storing of detection data while combining blockchain technology with machine learning classifiers to enable precise phishing website identification. While machine learning algorithms categorize websites according to phishing-related characteristics, blockchain smart contracts safely store detection results, performance metrics, and verification logs. This integration improves system dependability, stops result manipulation, and facilitates reliable decision-making in contemporary phishing website detection systems.