biglam/muninn-ww1-documents

Muninn WWI Documents (CEF Attestation Papers & War Diaries)

A tabular conversion of the document records in the Muninn Project's World War I Linked Open Data archive (https://rdf.muninn-project.org/). The Muninn Project is a research project that extracts structured data from digitized WWI-era archival documents.

The dataset covers 28,742 documents — 28,718 Canadian Expeditionary Force (CEF) attestation papers (enlistment forms of individual soldiers) and 24 Canadian unit war diaries — together with their 58,286 pages and 74,616 page-scan images held by Library and Archives Canada (LAC).

It was created for the BigLAM initiative (issue #52) by converting the Muninn RDF data dump into flat tables.

Configs

documents (default) — one row per archival document

pages — one row per scanned page

page_id, uri, document_uri (join key to documents.uri), title, page_number, image_uri, image_url, next_page_uri, previous_page_uri.

images — one row per scan image file

image_id, uri, title, image_url, format (GIF/JPEG), size_bytes, width_pixels, height_pixels, license, access_rights, date_retrieved.

Images are not bundled in this repository: image_url points to the original scans on Library and Archives Canada servers (data2.archives.ca, data2.collectionscanada.ca), which were still resolving as of July 2026.

Example uses

• —Document-type and country-of-origin classification of historical forms (e.g. identifying attestation papers from Canadian sources) from page scans.
• —Handwritten/typewritten text recognition (HTR/OCR) on early-20th-century forms, using the structured metadata (soldier name, birth date) as weak ground truth for the form fields.
• —Historical named-entity and record-linkage research on WWI personnel records.

How it was built

The Muninn SPARQL endpoint was offline when this dataset was created, so the conversion uses the project's published VoID data dump (muninn-Dump-Latest.nq, 2016, ~3.8M statements in SPARQL-XML form). The pipeline (scripts in conversion/):

1. 1.extract_quads.py — stream-parse the dump, keep statements with relevant predicates.
2. 2.build_tables.py — assemble Document → Page → Image join tables from the Muninn documents ontology (documents#contains, documents#depiction, documents#first_page, documents#back_page, …), resolve person names/birth dates from the organization and military ontologies, and write parquet.

Note: the 2016 dump itself was truncated by the publisher's export process, so this is a partial snapshot of the full Muninn triple store (which reported ~20M triples). All documents present are complete records.

Licensing

• —Metadata / RDF data: CC BY 3.0, as declared in the Muninn VoID description. Please attribute The Muninn Project.
• —Scanned images: Canadian Crown Copyright, expired — the Muninn data marks every document in this set PublicDomain. The scans are hosted by Library and Archives Canada; consult LAC terms for reuse of the image files themselves.
• —These are historical government records about (long-deceased) individuals; handle personal data (names, birth dates) in accordance with applicable research ethics norms.

Citation

@misc{muninn-ww1-documents,
  title   = {Muninn WWI Documents (CEF Attestation Papers and War Diaries)},
  author  = {{The Muninn Project} and {BigLAM initiative}},
  year    = {2026},
  note    = {Tabular conversion of the Muninn Project WWI Linked Open Data dump},
  url     = {https://rdf.muninn-project.org/}
}

Please also credit The Muninn Project (Rob Warren et al.) as the original data creator.