Primus-Seed: Cybersecurity Text Corpus from MITRE and Expert Sources

Primus-Seed is a cybersecurity text dataset compiled from reputable sources including MITRE, Wikipedia, and cybersecurity company websites, as well as manually collected Cyber Threat Intelligence (CTI). It was created by Trend Micro's AI Lab and was last updated on the Hugging Face platform in August 2025. The dataset includes at least 2,946 samples from cybersecurity blogs and news, comprising over 9.7 million tokens.

Use Cases

• Pretraining cybersecurity-specific language models based on the high-quality text corpus.
• Analyzing threat intelligence narratives based on expert-curated CTI reports.
• Benchmarking model performance on cybersecurity domain knowledge sourced from MITRE and Wikipedia.

Strengths

• Includes over 9.7 million tokens from cybersecurity blogs and news.
• Sources data from reputable entities like MITRE and Wikipedia.
• Contains manually collected Cyber Threat Intelligence (CTI) from threat experts.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the 'Web Crawl / Official Dump' category is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Trend Micro AI Lab (trendmicro-ailab).

Collection Method

Data crawled from MITRE, Wikipedia, and cybersecurity company websites, combined with manually collected CTI.

Freshness

Last updated 2025-08-08 08:01:28; freshness should be verified.