Develop and execute SPARQL competency questions #57

Affected: Documentation and test suites of DCAT-AP+ and its sub-profiles (currently only ChemDCAT-AP)
Severity: Evaluation gap - critical for demonstrating the architecture's value
Status: Not started

Problem: The current documentation and MTSR paper present instance data examples, but no systematic SPARQL competency questions are demonstrated against the generated RDF. This makes it difficult for adopters to assess whether the schema supports their query patterns, and for reviewers to evaluate the architecture's practical utility.

Proposed structure: three tiers:

Tier 1: DCAT-AP+ level (domain-agnostic queries, default vocabulary):

• "Find all datasets where any quantitative attribute has quantitykind:Temperature and value > 290"
• "Find all activities that evaluated entity X"
• "Find all datasets generated by activities involving device Y"

Tier 2: ChemDCAT-AP level (domain-specific queries, OBO/SIO vocabulary):

• "Find all chemical reactions that used a catalyst containing palladium and had yield > 80%"
• "Find all substance samples with InChIKey X"
• "Find all reactions where solvent was THF and temperature < -50°C"

Tier 3: Cross-profile (federated queries, projected vocabulary):

• Tier 1 queries executed against a KG containing both ChemDCAT-AP data (with projected DCAT-AP+ predicates) and data from another domain profile
• Demonstrates the structural interlingua claim: cross-domain discovery via shared graph shape
• can only be done once we have such another KG live

Deliverables: SPARQL .rq files in test suite, query results documented, referenced from project documentation. Tier 3 queries provide evidence for a potential journal publication extending the MTSR proceedings.