How IQ:NS Works
From documents to structured knowledge
Standards, regulations, and institutional rules exist as natural language in PDFs. Useful for people reading them cover to cover — less useful when you need a machine to reason about how Article 13 of the EU AI Act relates to a NIST control or an ISO clause.
IQ:NS models these concepts as formal ontologies — structured vocabularies where every term has a stable identifier, a definition grounded in its source, and explicit relationships to related concepts across frameworks.
The four layers
1. Concept modelling
Each standard is broken into its constituent concepts — obligations, controls, risk categories, roles — and represented as OWL classes with SKOS labels and definitions. Every concept traces back to its authoritative source.
2. Cross-framework alignment
Frameworks overlap constantly. GDPR Article 22 and EU AI Act Article 13 both address transparency. ISO 42001 and NIST AI RMF both define risk management processes. IQ:NS captures these alignments explicitly using skos:exactMatch, skos:broadMatch, and skos:relatedMatch.
One query shows you where frameworks converge, where they diverge, and where gaps exist.
3. Contextual profiling
Not every concept applies to every situation. IQ:NS supports profiling by jurisdiction, sector, and capability type — so you can query “what applies to a credit-scoring system in the EU?” and get a precise, deduplicated answer.
4. Continuous maintenance
Frameworks change. New standards arrive. The community maintains and extends the ontologies so the knowledge stays current. Every version is tracked — you can always see what changed and when.
How you use it
- Browse on GitHub — download the Turtle files, load into your triplestore
- Query via SPARQL — ask questions across frameworks programmatically
- Connect via MCP — let your AI agents reason over institutional knowledge directly
- Integrate with tools — plug into GRC platforms, data catalogs, CI/CD pipelines
What IQ:NS is not
- Not a policy engine — it doesn’t execute rules
- Not a GRC platform — it doesn’t run workflows or approvals
- Not a consultancy — you make the decisions; the ontologies provide the structure
IQ:NS is the semantic foundation those systems can build on.