Assessment of Vulnerability Severity in Power Monitoring System Based on Text Mining

Power Monitoring Systems (PMS) are essential for ensuring national energy security and societal stability. Remediation of vulnerabilities is labor-intensive due to the increasing severity of cybersecurity threats, which highlights the importance of effective severity assessments to prioritize. The standard method for assessing and prioritizing vulnerabilities uses the Common Vulnerability Scoring System (CVSS). However, the internal vulnerabilities specific to PMS do not have CVSS scores, making it difficult to assess their severity accurately. To remedy this problem, we present a text mining approach that achieves vulnerability severity predictions in PMS. The method conducts semantic analysis of vulnerability descriptions to extract key features and uses similarity analysis to rank vulnerabilities. Experimental results on real vulnerability data demonstrate the effectiveness of our approach in prioritizing vulnerabilities.