Abstract:
The world is witnessing a growing problem of cybersecurity as the number of threat actors that use backdoor networks and other communication tunnels to strategize, organize, and conduct advanced cyberattacks continues to increase. The security solutions are mostly reactive as they react to the occurrence of incidents when they have already occurred and caused damage to critical infrastructure.
This project proposes a novel automated threat intelligence system that is specifically designed to close this fundamental security gap by availing proactive predictions of any upcoming threats to any specific industry. This contrasts with other traditional security tools, which react to threats only after they occur and become actual attacks.
This system has a technical architecture that integrates advanced web crawling technology, Data Preprocessing pipeline specially tailored for this project, and artificial intelligence measures to have a threat intelligence technology fully automated. The solution uses custom-written crawling scripts that automatically search hidden network infrastructure with proper anonymization and access methods and systematically gather data with help of data processing pipeline that is associated to threat-related information on the hidden network sites which post data leaks. This curated data is then trained into a machine learning model that allows the system to automatically classify which specific industry can get affected in the upcoming time. The AI model uses natural language processing to gain context and extract useful information from the sites and is coupled with pattern recognition algorithms to determine specific information like URLs, dates, data sizes etc. And give proactive alert using dedicated Telegram channel.
