MALICIOUS DOMAIN NAME PREDICTION USING DEEP LEARNING MODEL

The identification of malicious domain names became increasingly crucial due to their rising threat to individuals, businesses, and digital infrastructure. Cybercriminals frequently used deceptive domain names to conduct phishing attacks, distribute malware, and compromise sensitive data. The lack of an efficient and automated detection system posed a significant challenge in mitigating these threats. This study addressed the problem by evaluating multiple machine learning algorithms for detecting malicious domain names, focusing on the Naïve Bayes Algorithm and the PageRank Algorithm due to their superior performance in binary and multi-class classification tasks. A dataset comprising malicious and benign domain names was collected from various cybersecurity repositories, including public threat intelligence databases and domain blacklists. The dataset underwent cleaning, preprocessing, and feature extraction before being used for training and testing. The system’s precision (0.482%), F1-score (0.514%), true positive rate (99.7%), and true negative rate (95.6%) further confirmed its reliability. For multi-class classification, the algorithm attained 93.0% accuracy, with strong precision, recall, and F1-score metrics, proving its effectiveness in identifying various categories of malicious domains. The system was developed using Python and Flask, ensuring scalability, responsiveness, and user-friendliness. A centralized database was implemented to manage domain records efficiently, reducing data redundancy and improving accessibility for users and administrators. The research followed the Agile software development methodology, enabling iterative improvements and flexibility in system enhancements. This study made a significant contribution to cybersecurity by enhancing malicious domain name detection. The findings demonstrated how machine learning models, combined with AI-driven ranking mechanisms, could effectively predict and mitigate cyber threats. This research laid the foundation for future advancements in web security, digital forensics, and automated threat intelligence systems.