Document is current

Article History

First Online: 7 November 2024

Declarations

: This research partially fulfills the requirements for the Ph.D. by publication at the University of Waikato in New Zealand. The Ministry of Business, Innovation, and Employment (MBIE) of the New Zealand Government funded this research under the Catalyst program for cybersecurity research. Z.Z. conducted the experiments presented in this paper independently without any financial incentives or Conflict of interest other than the MBIE funding and the prospected doctorial degree reward from the University of Waikato. All commercial entities mentioned in this research are present, solely due to the nature of CVE exposures. None of the aforementioned companies contacted the authors to influence the research results. There is no ethical approval needed for this research. The data involved in this research are publicly available from the National Vulnerability Database and the Common Vulnerabilities and Exposures database, crawled and stored in a public GitLab repository . The machine learning models and tools used in this research are open-source and publicly available in TensorFlow and Hugging Face. Z.Z. proposed the research methods, conducted the experiments, and composed the manuscript. V.K., the main supervisor of Z.Z., peer-reviewed weekly on the experiments with expertise in cybersecurity and administrated the research with aspects of funding, resources, and general guidance. B.P. offered machine learning insights and inspirations on machine learning methods in our weekly seminars. A.B. provided machine learning infrastructure through New Zealand Artificial Intelligence Institute and revised the manuscript.