CIRCL/vulnerability-scores

This dataset, CIRCL/vulnerability-scores, comprises over 600,000 real-world vulnerabilities used to train and evaluate VLAI, a transformer-based model designed to predict software vulnerability severity levels directly from text descriptions, enabling faster and more consistent triage.

The dataset is presented in the paper VLAI: A RoBERTa-Based Model for Automated Vulnerability Severity Classification.

Project page: https://vulnerability.circl.lu Associated code: https://github.com/vulnerability-lookup/ML-Gateway

Sources of the data

• CVE Program (enriched with data from vulnrichment and Fraunhofer FKIE)
• GitHub Security Advisories
• PySec advisories
• CSAF Red Hat
• CSAF Cisco
• CSAF CISA

Extracted from the database of Vulnerability-Lookup.
Dumps of the data are available here.

Query with datasets

import json
from datasets import load_dataset

dataset = load_dataset("CIRCL/vulnerability-scores")

vulnerabilities = ["CVE-2012-2339", "RHSA-2023:5964", "GHSA-7chm-34j8-4f22", "PYSEC-2024-225"]

filtered_entries = dataset.filter(lambda elem: elem["id"] in vulnerabilities)

for entry in filtered_entries["train"]:
    print(json.dumps(entry, indent=4))