APPLICATION OF ARTIFICIAL INTELLIGENCE FOR MONITORING DARK WEB ACTIVITIES

Abstract

The Dark Web is a hidden part of the internet that attracts perpetrators due to its anonymity. It is used for cybercrime, illegal trading, and by terrorist groups to exchange stolen data and personal information. Current techniques for combating and identifying these concerns are weak. The primary focus of this study is on the application of artificial intelligence to monitor the Dark Web. The solution mitigates monitoring deficiencies to facilitate the timely identification of prospective threats, including data breaches and cybercrimes. The six forms of activities prioritised include cyberterrorism, terrorist activities, weapons trading, drug trafficking, human trafficking, and regular activities. The tasks carried out in this study include crawling relevant data, training machine learning models using an ensemble voting approach, and implementing a JavaScript-powered crawling engine. The AI model is tested and evaluated using data collected from various Onion sites that include both legal and illegal content. The approach used in this study gives 97% accuracy and a macro average precision of 98%. The macro average for recall and f1-score is 97%. The precision, recall and F1-Score all have the same weighted average of 97%.