Skip to main content

DPCH11 — Semantically Aligned

“Aligned to Enterprise Vocabulary / Ontology”

What DPCH11 is really asserting​

DPCH11 is not asserting that:

“Column names look reasonable” or “someone used similar words.”

It is asserting that:

The Data Product’s meaning is formally aligned to shared enterprise and domain semantics so that its data can be interpreted, combined, and reasoned about consistently across products, domains, and consumers (human and machine).

This is about shared meaning, not naming hygiene.

The Essence (HDIP + Data Mesh Interpretation)​

A Data Product is semantically aligned if and only if:

  1. Its key concepts map to canonical enterprise or domain definitions
  2. Meaning is explicit and machine-referenceable
  3. The product can interoperate with other products without semantic guesswork

If combining two products requires:

  • meetings to reconcile meaning,
  • manual mapping spreadsheets,
  • tribal knowledge of “what this field really means”,

then DPCH11 is not met, even if schemas look similar.

Positive Criteria — When DPCH11 is met​

DPCH11 is met when all of the following are true:

1. Business concepts are anchored to shared definitions​

The Data Product:

  • references enterprise or domain glossary terms
  • references enterprise or domain conceptual ontology
  • uses canonical identifiers where defined
  • clearly states alignment or intentional deviation

Example:

  • payment_id explicitly aligned to Enterprise Payment Identifier
  • settlement_date mapped to agreed time semantics

2. Semantic alignment is explicit, not inferred​

Alignment is:

  • declared in product metadata
  • visible in catalog / product page
  • consumable by machines (not just prose)

Implicit similarity (“looks the same”) does not count.

3. Semantic decisions are stable and governed​

Once aligned:

  • changes follow versioned evolution
  • breaking semantic changes are explicit
  • consumers are informed through product lifecycle signals

This preserves trust and reuse.

Negative Criteria — When DPCH11 is not met​

DPCH11 is not met if any of the following are true:

❌ Semantics are local to the product​

Examples:

  • definitions only make sense within one team
  • field meanings differ subtly from enterprise usage
  • no mapping to shared vocabulary exists

This leads to semantic fragmentation.

❌ Alignment is “documented” but not structured​

Examples:

  • free-text descriptions saying “same as X”
  • slides explaining differences
  • Confluence pages with mappings

Humans might cope; machines cannot.

❌ Product deliberately ignores enterprise semantics​

Examples:

  • “We didn’t like the canonical model”
  • “We created our own version for speed”
  • “We’ll align later”

This breaks interoperability and long-term reuse.

Edge Cases (Important Guidance for Agents)​

Case 1: “Enterprise vocabulary exists but product not aligned”​

❌ Not met

Rationale:

  • availability of ontology ≠ usage
  • alignment is a conscious act

Case 2: “Partial alignment with clear declaration”​

⚠️ Partial

Rationale:

  • some concepts aligned
  • others explicitly marked as local extensions
  • acceptable transitional state

Case 3: “Full alignment or explicit, versioned deviation”​

✅ Met

Rationale:

  • shared meaning preserved
  • deviations are intentional and visible

Evidence Signals an Agent Should Look For​

Authoritative evidence:

  • glossary / ontology references
  • canonical identifiers in schemas
  • semantic mapping artifacts linked to product

Supporting evidence:

  • enterprise data model links
  • cross-product join examples that require no reinterpretation

Red flags:

  • “business knows what this means”
  • duplicate concepts with different names
  • undocumented semantic differences

How an AI Agent Should Decide​

Decision rule (simplified):

If the meaning of the Data Product’s core concepts cannot be reliably interpreted and combined with other products using shared enterprise semantics, DPCH11 is not met.

Why DPCH11 Is Non-Negotiable​

Without DPCH11:

  • reuse becomes dangerous
  • analytics produce conflicting results
  • AI agents hallucinate meaning
  • Data Mesh degenerates into semantic silos

DPCH11 is what allows federation without fragmentation.

Canonical Statement​

DPCH11 is satisfied only when a Data Product’s meaning is explicitly aligned to shared enterprise or domain vocabularies and conceptual ontologies, enabling consistent interpretation and safe interoperability across products and consumers.