5G CVE Dataset

This repository contains the dataset for the pubblication Bene, P., Bernardini, A., Sagratella, L., Maunero, N., & Settembre, M. (2025, September). Optimizing Local LLM Deployment for 5G CVE Classification Avoiding External Data Exposure. In 2025 IEEE Conference on Communications and Network Security (CNS). IEEE. This dataset contains 136 manually annotated CVE records sourced from the NIST National Vulnerability Database (NVD) in the period 2014-2024, with a specific focus on accurate classification of vulnerabilities truly correlated to 5G network technologies. The CVE entries are annotated with two labels: 5G: Vulnerabilities relevant to 5G core network functions, RAN components, or 5G-specific protocols. Non-5G: Vulnerabilities related to general networks, applications, or infrastructure not specific to 5G. This balanced dataset (68 5G entries and 68 Non 5G entries) can be used as a ground truth for supervised learning algorithms.