Development of a Machine Learning-Based Cyber Threat Intelligence Dashboard System for Strategic Operations Centre

Abstract

Cyber Threat Intelligence (CTI) has become an essential element in the toolkit of Cybersecurity experts. In recent years, the significance of CTI has grown exponentially due to the increasing sophistication and frequency of cyber attacks. The incorporation of machine learning methodologies into CTI systems represents a substantial advancement in the domain. Conventional rule-based systems frequently fall short in identifying emerging threats and adjusting to the swiftly evolving strategies employed by cybercriminals. This paper presents a systematic appraisal of CTI dashboard systems that incorporate machine learning techniques to enhance strategic cybersecurity operations, which provide a user-friendly platform for real-time threat detection, analysis, and visualisation. At the core of this study is the utilisation of Gradient Boosting Trees (GBT) as the primary machine learning algorithm for threat detection and classification. The research only focused on the detection, analysis, and presentation of threat intelligence, leaving the specific response strategies at the discretion of the organisation implementing the system. The CTI dashboard system, which is the result of this work, showed strong performance, with a precision of 99.6%, a recall of 99.5%, and an F1-score of 99.97%. The system also showed an average response time of 3 minutes and 12 seconds, demonstrating its effectiveness in delivering timely and accurate threat intelligence.