Towards autonomous network security using knowledge graphs



Journal Title

Journal ISSN

Volume Title



The rapid growth of cyber threats in recent years has increased the demand for effective and efficient methods to map and analyze computer networks. In this thesis, we present a novel autonomous system for deploying agents through computer networks and collecting host information to create a comprehensive knowledge graph. The knowledge graph captures details about the installed software, operating systems, and various other aspects of the host, providing valuable insights for network analysis and security purposes. Our evaluation focuses on coverage, scalability, and efficiency, with experiments conducted on both real-world and simulated network environments of varying sizes and configurations. We assess the system's performance in open, semi-open, and hardened configurations. Ultimately, this comprehensive evaluation aims to validate the system's capabilities and demonstrate its potential impact on network analysis and security. The evaluation results demonstrate the system's effectiveness in mapping networks and collecting accurate host information while maintaining a low impact on the network. However, the evaluation also reveals areas for improvement, particularly regarding environments with strict security measures and the scalability issues of hosting an API on a single server. Based on our findings, we propose potential future improvements, such as incorporating machine learning and AI techniques to enhance the system's adaptability and resilience. Overall, our autonomous system offers a promising approach to network analysis and security, laying the foundation for further research and innovation in this field.



autonomous, network security, knowledge graph, cybersecurity