Modeling for Identifying Attack Techniques Based on Semantic Vulnerability Analysis

As attackers' techniques and tactics targeting Industrial Control Systems (ICS) become increasingly diverse and sophisticated, attacker-centric security validation technologies are essential for detecting and mitigating such attacks. Study has been conducted to identify attack techniques based on information about vulnerabilities and the root causes of cyberattacks. However, existing studies rely on word embeddings of textual information, which have limitations in understanding the overall context of sentences. This study proposes a modeling approach to identify attack techniques based on the semantic analysis of vulnerability information to resolve these limitations. This approach demonstrates specific attack paths from vulnerability information through a comparative analysis with expert evaluations. Furthermore, it suggests the potential to identify detection and mitigation measures for vulnerabilities.