LLM based approach to CPE matching from asset inventory

In the context of Vulnerability Management, the identification of Common Platform Enumerations (CPEs) is crucial. This study proposes a novel solution to the specific challenge of extracting CPEs from asset inventories by using Large Language Models (LLMs) in order to associate each entry of an asset inventory to its corresponding CPE. The study also addresses all the secondary aspects related to the use of LLMs with the purpose of providing a comprehensive overview of the solution and its implications.