With a growing amount of information security incidents around the world, organizationsneed to manage information security more efficiently. A way to enable organizations to improve their information security management is to utilize decision support systems in information security. Previous studies has presented promising capabilities in machine learning models for analysis of security vulnerabilities with the industry standard Common Vulnerability Scoring System 2.0. These studies hashowever used the older version of the scoring system, and not in all cases fully automated the entire analysis process. This research conducts an experiment which indicates that the newer scoring system, Common Vulnerability Scoring System 3.0 is possible to automate with machine learning models. The machine learning models in this study perform similarly and in some cases slightly better than the previous studies. This study presents the possibility of a completely automated scoring system, the study presents a high positive correlation of 0.7 with classifications from the recognized information security database NVD which publishes information security analyses for vulnerabilities in systems.