Skip to main content

AIPS DPP Specification Downloads

Purpose

This page provides direct download links to all machine-readable and human-readable artifacts that define the AIPS DPP Profile (v0.1).

The AIPS DPP Profile extends the BPS DPP Core to describe AI Products with additional model, evaluation, and risk metadata, while preserving compatibility with the base DPP verification and trust framework.

All artifacts listed below are publicly accessible under the versioned namespace:


[https://kivanura.org/spec/aips/dpp/0.1/](https://kivanura.org/spec/aips/dpp/0.1/)


1. Machine-Readable Artifacts

FileTypePurpose
aips-dpp-0.1.context.jsonldJSON-LD ContextDefines AI-specific DPP vocabulary (e.g., modelCard, eval, risk, policy).
aips-dpp-0.1.schema.jsonJSON SchemaMachine-validated structure for AIPS DPP instances.
aips-dpp-0.1.shacl.ttlSHACL ShapesSemantic validation rules compatible with RDF/Linked Data tools.
aips-dpp-0.1.ttlRDF VocabularyOntological representation of the AIPS DPP Profile.
aips-dpp-0.1.openapi.yamlOpenAPI ContractDefines API endpoints for retrieving Lite and Full AIPS DPPs.

Each of these files aligns structurally with the corresponding BPS DPP Core 0.1 artifacts and can be validated or combined programmatically.


2. Human-Readable Guides

FileTypeDescription
aips-dpp-overview.mdMarkdownOverview of scope, composition, and relationship with BPS Core.
aips-dpp-implementation-guide.mdMarkdownStep-by-step guidance for producers issuing AI Product DPPs.
aips-dpp-consumers-guide.mdMarkdownGuidance for consumers on reading, validating, and verifying AIPS DPPs.
aips-dpp-examples.mdMarkdownRealistic Lite and Full DPP examples for reference.

These documents complement the machine-readable files by describing usage, validation, and implementation best practices.


3. Example Instances

FileTypeDescription
fraud-detector-lite.example.jsonldExample InstanceAIPS Lite Passport for a FraudDetector AI model.
fraud-detector-full.example.jsonldExample InstanceAIPS Full Passport for the same model, including ByRef sections.
vision-classifier-lite.example.jsonldExample InstanceCompact DPP for an image-classification AI product.

All example instances are syntactically valid according to the AIPS DPP schema and SHACL constraints.


4. Alignment and References

FilePurpose
alignment.mdDescribes semantic alignment between AIPS DPP and BPS DPP Core vocabularies.
security-privacy.mdOutlines redaction, privacy, and signature-handling expectations.
versioning-policy.mdExplains the version management model and lifecycle semantics.
CHANGELOG.mdTracks version history, updates, and backward-compatibility notes.

5. Namespace Summary

PrefixIRI
aipsdpp:https://kivanura.org/spec/aips/dpp/0.1/
bpsdpp:https://kivanura.org/spec/bps/dpp/0.1/
xsd:http://www.w3.org/2001/XMLSchema#
rdf:http://www.w3.org/1999/02/22-rdf-syntax-ns#
sh:http://www.w3.org/ns/shacl#

Namespaces ensure that every AIPS DPP instance can interoperate with the broader BPS ecosystem and RDF/Linked Data tooling.


6. Usage and Validation Notes

  • All JSON-LD and Turtle artifacts can be validated using:

    • pyld or rdf-validate-shacl for semantic checks.
    • ajv or similar libraries for JSON Schema compliance.
  • Producers must include both contexts in AIPS DPP instances:

    "@context": [
    "https://kivanura.org/spec/bps/dpp/0.1/bps-dpp-core-0.1.context.jsonld",
    "https://kivanura.org/spec/aips/dpp/0.1/aips-dpp-0.1.context.jsonld"
    ]

* Consumers should verify signatures and content hashes per the BPS DPP Core validation process.

---

## 7. Version and Governance

* **Version:** 0.1 (Initial Release)
* **Namespace:** `https://kivanura.org/spec/aips/dpp/0.1/`
* **Maintainer:** Kivanura Foundation
* **License:** CC BY-SA 4.0
* **Change Policy:** Minor versions maintain backward compatibility; major versions introduce schema or vocabulary extensions.

Future versions (e.g., `0.2`, `1.0`) will remain linked under the parent path:

```
https://kivanura.org/spec/aips/dpp/
```

---

## 8. Summary

The AIPS DPP Profile provides a structured, verifiable, and interoperable mechanism for representing the **digital trust layer of AI Products**.
By combining the BPS DPP Core with domain-specific extensions, it enables transparent model governance and reproducibility across the entire AI product lifecycle.

---