Skip to content

Vocabularies and claims

Shared vocabularies give ASB slots a meaning outside the schema; claim and identifier formats carry evidence and releases beyond it.

Term mappings

Each slot is aligned with external vocabularies through LinkML mappings: exact_mappings for terms of the same meaning and close_mappings for terms of similar meaning1. The seven LinkML files hold 72 exact and 120 close mappings.

Vocabulary Description Mappings Examples
schema.org General web vocabulary for things, creative works and datasets2 92 title to schema:name, text to schema:text, registry_iri to schema:url
PROV-O W3C ontology of provenance: entities, activities, agents3 45 source_doi to prov:wasDerivedFrom, generated_at to prov:generatedAtTime
Dublin Core terms Bibliographic metadata terms 22 snippet_id to dcterms:identifier, collections to dcterms:hasPart
EDAM Ontology of data analysis in the sciences: topics, operations, data, formats4 21 edam_operation, edam_topics, domain_topics hold EDAM IRIs
P-Plan Extension of PROV-O for plans, steps and variables5 11 evaluation_strategy close to pplan:Plan
RO-Crate Research object packaging 1 ro_crate_version

Slots also carry an importance annotation: must_have, should_have or nice_to_have. Mappings are annotations; the generated JSON-LD context sets @vocab to the ASB Schema namespace, so a JSON-LD processor expands keys there (see Conformance). Capsule crates use schema.org, PROV-O and P-Plan types directly.

bio.tools, the registry of bioinformatics software, identifies tools by biotoolsID; ASBTool reserves bio_tools_id for that link (feasible). A release gate can check EDAM IRIs against a pinned EDAM release6.

SSSOM

The Simple Standard for Sharing Ontological Mappings (SSSOM) exchanges mappings between semantic entities, most often as a TSV file with a metadata block7. asb export-sssom writes skill-to-EDAM mappings to mappings.sssom.tsv.

existing

Claims and evidence

A card types the evidence for its task: EvidenceSnippet requires snippet_id, source_path, reason and text, so each verbatim passage travels with the reason it supports the card.

{
  "snippet_id": "snip_001",
  "source_path": "paper/methods.md",
  "reason": "States the retention-time window that the task uses to align features across samples.",
  "text": "Features were aligned across samples within a retention-time window of 0.1 min."
}

SCHEMA=$(python -c "from asb_schema import get_schema_path; print(get_schema_path('asb_card'))")
linkml-validate --schema "$SCHEMA" --target-class EvidenceSnippet snippet.json
# No issues found

Two card slots link a card to the claim ledger it was built from: linked_result_ids (the results the card grades) and section_provenance (source sections taken from the anchors of the linked claims).

Bridge Description Status
Claim ledger to EvidenceSnippet Each claim keeps a verbatim excerpt and a character span; an adapter turns a claim into a snippet existing
Claim ledger to ASTRA findings Claims become insights with quoted evidence in astra.yaml existing
Claim records to nanopublications One nanopubs/claim_NNN.jsonld per claim, with head, assertion, provenance and publication-information graphs8 existing
EvidenceSnippet to ASTRA Evidence See ASTRA feasible

A slot on EvidenceSnippet holding the IRI of the supporting claim would connect card evidence to nanopublications built from the ledger.

Archive and identifiers

Format Use in ASB Status
w3id.org Persistent namespaces for schemas, terms, collections and the registry9 existing
Zenodo and DataCite A deposit per collection version, prepared in .zenodo.json; the DOI in doi and zenodo_doi. Zenodo registers one DOI per version and a concept DOI for all versions10 existing
Citation File Format CITATION.cff in this repository and in each released collection, which lists the source papers under references11 existing
CodeMeta Software metadata in JSON-LD, generated from CITATION.cff with cffconvert1213 feasible
Croissant JSON-LD metadata for machine-learning datasets; a benchmark layer could be a cr:FileSet of cards with a cr:RecordSet of card slots14 feasible

See also

  1. LinkML, "URIs and mappings". https://linkml.io/linkml/schemas/uris-and-mappings.html ↩

  2. schema.org, "About schema.org". https://schema.org/docs/about.html ↩

  3. W3C, "PROV-O: The PROV Ontology". https://www.w3.org/TR/prov-o/ ↩

  4. EDAM, "edamontology" repository. https://github.com/edamontology/edamontology ↩

  5. OPMW, "The P-Plan Ontology". https://www.opmw.org/model/p-plan/ ↩

  6. EDAM, "EDAM.tsv". https://edamontology.org/EDAM.tsv ↩

  7. Mapping Commons, "SSSOM: Simple Standard for Sharing Ontological Mappings". https://mapping-commons.github.io/sssom/dev/ ↩

  8. Nanopublications, "Nanopublication Guidelines" (working draft). https://nanopub.net/guidelines/working_draft/ ↩

  9. w3id.org, "Permanent Identifiers for the Web". https://w3id.org/ ↩

  10. Zenodo, "Versioning". https://zenodo.org/help/versioning ↩

  11. Citation File Format, "Citation File Format". https://citation-file-format.github.io/ ↩

  12. CodeMeta, "CodeMeta". https://codemeta.github.io/ ↩

  13. cffconvert, "cffconvert" repository. https://github.com/citation-file-format/cffconvert ↩

  14. MLCommons, "Croissant format specification". https://docs.mlcommons.org/croissant/docs/croissant-spec.html ↩