The Advancement of Knowledge Graphs in Cybersecurity: A Comprehensive Overview

With the increasing complexity of artificial intelligence technology and network environments, cybersecurity is facing massive and complex data. Knowledge graphs have the potential to aggregate, represent, manage, and reason with this knowledge. Therefore, applying knowledge graphs to cybersecurity can help to characterize and present security situations, support security decision-making, and predict warnings. Over the past two decades, research on knowledge graphs for cybersecurity has received growing attention in data processing, construction, and visualization. This review provides a comprehensive comparative analysis of key technologies and application scenarios of cybersecurity knowledge graphs. Firstly, basic concepts of knowledge graphs and cybersecurity knowledge graphs are outlined, and the required datasets for their construction are compared and analyzed from both general-purpose and specialized perspectives. On this basis, a framework for building cybersecurity knowledge graphs is summarized, and key techniques for building cybersecurity knowledge graphs, including ontology construction, information extraction, and knowledge reasoning, are detailed. Finally, application scenarios of knowledge graphs in the field of cybersecurity are sorted out from the perspective of application objectives. The challenges knowledge graphs face and future development trends in this field are also pointed out.