stefdoerr/cidoc-crm-corpus
CIDOC CRM SIG corpus — derived artifacts The built corpus for cidoc-crm-mcp: 26 years of CIDOC CRM Special Interest Group mailing list, its issue register and meeting minutes, cleaned, threaded, and indexed for BM25 and vector search. These are derived artifacts, not the source. They exist because the code repository cannot carry them: ~876MB, git-ignored, rebuilt from a 143MB mbox that is distributed separately. Fetching uv run python build.py fetch… See the full description on the dataset page: https://huggingface.co/datasets/stefdoerr/cidoc-crm-corpus.
CIDOC CRM SIG corpus — derived artifacts
The built corpus for cidoc-crm-mcp: 26 years of CIDOC CRM Special Interest Group mailing list, its issue register and meeting minutes, cleaned, threaded, and indexed for BM25 and vector search.
These are derived artifacts, not the source. They exist because the code repository cannot carry them: ~876MB, git-ignored, rebuilt from a 143MB mbox that is distributed separately.
Fetching
uv run python build.py fetch # everything uv run python build.py fetch --no-vectors # 91MB, skips the vectors
Without the vectors you still get full-text search over the whole archive and you do not need torch or an embedding model.
Contents
Provenance
Each vector store records the embedding model it was built with and a source_sha256 of the input it was built from. build.py fetch checks these, because a store embedded with one model and queried with another returns confident nonsense.
Licence
The mailing list is published by the CIDOC CRM SIG; posts remain their authors'. This is a derived index redistributed for research use. Contact the SIG for reuse beyond that.
