Relation extraction techniques in cyber threat intelligence

Cyber Threat Intelligence (CTI) provides a structured and interconnected model for threat information through Cybersecurity Knowledge Graphs. This allows researchers and practitioners to represent and organize complex relationships and entities in a more coherent form. Above all, the discovery of hidden relationships between different CTI entities, such as threat actors, malware, infrastructure, and attacks, is becoming a crucial task in this domain, facilitating proactive defense measures and helping to identify Tactics, Techniques, and Procedures (TTPs) employed by malicious parties. In this paper, we provide a Systematization of Knowledge (SoK) to analyze the existing literature and give insights into the important CTI task of Relation Extraction. In particular, we design a categorization of the relations used in CTI; we analyze the techniques employed for their extraction, the emerging trends and open issues in this context, and the main future directions. This work provides a novel and fresh perspective that can help the reader understand how relationships among entities can be schematized to provide a better view of the cyber threat landscape.