CVE
1. Utilizing embedding techniques to compute vectors for each tactic/technique in the Mitre ATT&CK matrix. For each CVE entry, peforming vectorization and calculating cosine similarity with the vectors of each tactic/technique to identify the top similar tactics.
2. Leveraging the analytical capabilities of Large Language Model (LLM) to conduct semantic and contextual analysis of host information and CVE configurations. This facilitates the identification of potential vulnerabilities and security risks through comprehensive analysis and correlation.