Published

Semantic Model

vr.wm-dat-015 · wm-dat-015-semantic-model

Describe the business meaning of datasets through a governed, versioned composition of entity roles, relationships, metric bindings and dimension bindings.

World Models Information and virtual systems INF.DAT.SEM

Bundle → Layer → Finding → Questions Classifiers What it is Why it exists Distinguishing features What robots and AI may and may not do Moral aspects Owners Links to other meta-models Identity and identifiers Direct properties Recognition Capabilities and actions Hazards and failure modes Standards and interfaces Context of use Sources

Bundle → Layer → Finding → Questions Filled

6 bundles · 12 layers · 12 findings · 48 questions

Identity and meaning Identify the definition and the business vocabulary it applies.

Definition mandate

Proposed root record identifies one semantic definition, its intended use and authoritative steward. A composed view remains one governed entity; referenced definitions keep their own identities.

Semantic definition identity and remit

Proposed root record identifies one semantic definition, its intended use and authoritative steward. A composed view remains one governed entity; referenced definitions keep their own identities.

Which master identifier denotes this semantic definition across revisions? identity
Which business questions and dataset roles fall within this model? definition
Which accountable role can approve this definition for its stated use? ownership
Is this revision conceptual, bound to data, or retired from use? classification

Business vocabulary

Bind local business terms to concept references with language and context. Similar names do not establish shared meaning, and mapping strength remains explicit.

Contextual term and concept bindings

Bind local business terms to concept references with language and context. Similar names do not establish shared meaning, and mapping strength remains explicit.

Which concept and definition does each local business term denote? definition
Which labels and language tags distinguish synonyms from homonyms? classification
What mapping strength and evidence justify each external concept alignment? interoperability
How is an unresolved or disputed business term represented without silently selecting a meaning? exception

Entities and relationships Declare the semantic units represented and the paths allowed between them.

Semantic units

Declare what one record or observation represents in each dataset role. Distinguish a business entity from an observation about it and from its storage key.

Entity role and observation grain

Declare what one record or observation represents in each dataset role. Distinguish a business entity from an observation about it and from its storage key.

What distinguishes one business entity from one observation in each role? identity
At what grain is each selected dataset role interpreted? composition
Which key and uniqueness assumptions are declared and which have supporting evidence? constraint
How are duplicate observations, absent keys and changing entity identifiers handled? exception

Analytical paths

Keep the semantic predicate separate from its physical join binding. Proposed path restrictions require explicit cardinality and grain evidence; executable join safety is not claimed.

Relationship meaning and join admissibility

Keep the semantic predicate separate from its physical join binding. Proposed path restrictions require explicit cardinality and grain evidence; executable join safety is not claimed.

What predicate, direction and role distinguish the endpoints of each relationship? relationship
Which endpoint cardinalities and temporal conditions must hold for an analytical path? constraint
How will reviewers detect fan-out, unmatched rows and ambiguous paths before admitting a binding? quality
What restriction applies when a many-to-many path lacks an approved allocation rule? exception

Analytical composition Bind independently governed metrics and dimensions into a meaningful view.

Metric bindings

Select metric definitions and map their input roles to the model. Formula changes belong to a metric revision; this definition records contextual admissibility and loss of meaning.

Contextual metric selection

Select metric definitions and map their input roles to the model. Formula changes belong to a metric revision; this definition records contextual admissibility and loss of meaning.

Which metric identifiers and revisions are exposed by this semantic model? composition
Which source roles, units and grain satisfy each metric input contract? measurement
Which dimensions and aggregation levels are allowed or prohibited for each selected metric? constraint
What happens if a metric dependency changes or its required input becomes unavailable? exception

Dimension roles

Assign local analytical roles and eligible levels to separately maintained dimensions. Hierarchy membership and semantic similarity do not prove that a measure can be summed.

Dimension and hierarchy selection

Assign local analytical roles and eligible levels to separately maintained dimensions. Hierarchy membership and semantic similarity do not prove that a measure can be summed.

Which dimension and hierarchy revisions serve each analytical role? composition
Which levels or code subsets are applicable to this view? classification
Which dimension validity rules apply to historical and current views? temporal
How are unknown members, overlapping groups and non-additive roll-ups exposed? exception

Value and time interpretation Make context-dependent interpretation explicit before combining observations.

Value interpretation

Record semantic interpretation for bound values, with conversion and missing-value rules delegated to approved definitions. Unknown, absent, suppressed, not applicable and zero are distinct proposed statuses.

Units, missingness and comparability

Which unit, scale and representation apply to each value role? measurement
How are zero, missing, suppressed and not-applicable values distinguished? definition
Which conversions or comparisons require a separately approved rule? constraint
What evidence supports comparability across sources or changed measurement methods? quality

Temporal context

Separate observation period, event time, processing time and definition validity. RFC 3339 timestamps do not encode a complete reporting calendar or a fiscal period.

Analytical time and applicability

Separate observation period, event time, processing time and definition validity. RFC 3339 timestamps do not encode a complete reporting calendar or a fiscal period.

Which time role governs each observation and each metric binding? temporal
What timezone, calendar and boundary convention determine period membership? constraint
Which revision applies to an as-of query after a retroactive correction? lifecycle
How are late observations, incomplete periods and unknown time precision declared? exception

Bindings and acceptance Connect the semantic definition to representations and scoped test evidence.

Source mappings

Reference approved mappings between conceptual roles and source fields or expressions. A mapping describes intended correspondence; execution and transformation lineage remain external.

Representation binding and semantic loss

Reference approved mappings between conceptual roles and source fields or expressions. A mapping describes intended correspondence; execution and transformation lineage remain external.

Which schema revisions and field paths implement each semantic role? interoperability
Who supplied the mapping and which evidence supports its meaning? provenance
Which casts, filters or transformations may change the interpreted population or value? constraint
How is an unresolved field, changed schema or lossy export reported? exception

Evidence and acceptance

Record acceptance criteria and links to supplied reports for this revision and binding set. Distinguish report failure, nonconformance, absence of evidence and accepted limitations.

Scoped semantic acceptance evidence

Record acceptance criteria and links to supplied reports for this revision and binding set. Distinguish report failure, nonconformance, absence of evidence and accepted limitations.

Which acceptance criteria test meaning, grain and the selected metric-dimension combinations? validation
Which exact model, source snapshot and validator configuration produced a supplied report? evidence
Which targets or paths were untested even if a report says the selected inputs conform? quality
Who accepts residual limitations and what prevents an untested binding from being labelled ready? decision

Revision and use governance Govern the definition and its disclosures without operating underlying systems.

Change and provenance

Maintain the semantic definition revision chain and dependency pins. Proposed change reviews classify meaning changes and notify affected roles through a separate authorized process.

Revision lineage and dependency impact

Maintain the semantic definition revision chain and dependency pins. Proposed change reviews classify meaning changes and notify affected roles through a separate authorized process.

Which prior revision and evidence explain each semantic change? provenance
Which review and approval evidence authorizes activation or retirement of a revision? lifecycle
Which dependent views or bindings are incompatible with a change in grain, term or component revision? constraint
What record disposition preserves necessary references while removing restricted payloads when required? retention

Policy context

Reference policies applying to definitions, mappings and evidence. Public vocabulary does not grant access to underlying data or authorize inference from combined sensitive dimensions.

Use restrictions and disclosure boundary

Reference policies applying to definitions, mappings and evidence. Public vocabulary does not grant access to underlying data or authorize inference from combined sensitive dimensions.

Who may inspect the definition, mappings and linked evidence at each structural scope? access
Which combinations of exposed roles or measures require restricted disclosure review? privacy
Which external authority decides actual data access and permitted execution? authority
How is a time-limited policy exception recorded without silently widening the default access rule? exception

Classifiers Filled

Family: World Models
Category: Information and virtual systems
Entry kind: entity
Navigation path: NAV.INF.DAT.SEM
Domain: INF.DAT.SEM
Industry: Cross-industry
Tags: semanticmodelinf.dat.sem

What it is Filled

One semantic definition with stable identity and separately identified revisions. It specifies how selected datasets are interpreted for a stated use, including conceptual entities, grain, relationships and bindings to separately governed metrics and dimensions. It may be drafted before data is available. Storage, query execution and physical schemas are external. All structure and operations below are proposed local design, informed by open specifications rather than claimed universal requirements.

In scope

Model-specific scope, term bindings, entity roles and declared populations or observation grain
Relationship meaning, admissible analytical paths, metric and dimension bindings with contextual restrictions
Interpretation of time, units and missingness; physical binding references and scoped acceptance evidence
Provenance, release applicability and policy references for the semantic definition itself

Out of scope

Dataset contents, distributions and storage lifecycle; physical schema and producer-consumer agreement masters
Independent metric formula authoring, dimension member and hierarchy maintenance, ontology or glossary lifecycle
Query compilation and execution, actual authorization, data transformation, runtime lineage capture and monitoring
A universal ontology, commercial analytical platform configuration or a claim of executable standard conformance

Why it exists Filled

Describe the business meaning of datasets through a governed, versioned composition of entity roles, relationships, metric bindings and dimension bindings.

Distinguishing features Derived, awaiting review

Unlike WM-DAT-004 Data Schema / Data Contract: The registry parent is a taxonomy lead. Align the specialized semantic composition to an external schema or contract; do not infer a producer-consumer agreement from the existence of business meaning. Generic schema constraints and contractual obligations stay with that master. No executable inheritance or mandatory contract instance is ratified.
Unlike WM-DAT-001 Dataset: A semantic definition can describe several datasets or an intended dataset role. A source binding references the selected asset and revision, while records, access enforcement and asset lifecycle remain external.
Unlike WM-DAT-013 Metric Definition: This model selects a metric revision and binds its inputs and permitted dimensions. Formula, generic aggregation semantics and metric ownership remain in the metric master. A local override that changes meaning requires a separately governed metric revision.
Unlike WM-DAT-014 Dimension and Hierarchy: This model assigns analytical roles, permitted levels and context restrictions to pinned dimension references. It does not maintain member lists, hierarchy edges or their lifecycle.
Unlike WM-KNW-001 Model / Ontology and WM-DAT-011 Knowledge Graph: Term and ontology references constrain interpretation but do not turn this definition into a universal ontology or the graph of instance assertions. Graph publication is a projection, not the aggregate root.
Unlike WM-DAT-005 Data Pipeline and external evaluators: Mapping and assessment references document intended interpretation and supplied evidence. They do not own data transformation runs, query plans, metric calculation, evaluator internals or enforcement state.

Note: Derived from boundary notes against neighbouring models.

What robots and AI may and may not do Derived, awaiting review

Must not

Deny access to restricted mapping and evidence details until purpose and role scope are authorized.
Require a referenced approval with purpose, scope, expiry and revocation route; an exception cannot silently propagate to external data.

May

Prepare semantic revision: Proposed local operation, not implemented: prepare a candidate revision with explicit unresolved terms and dataset roles.
Propose component bindings: Proposed local operation, not implemented: bind supplied metric, dimension and dataset revisions to local roles.
Inspect declaration completeness: Proposed local operation, not implemented: check the proposed record for unresolved references and missing admission evidence.
Attach supplied acceptance evidence: Proposed local operation, not implemented: index externally produced results against an exact revision and scope.
Prepare semantic change assessment: Proposed local operation, not implemented: compare supplied revision declarations and dependency pins.
Record authorized retirement: Proposed local operation, not implemented: mark a definition revision unavailable for new use after a supplied decision.

Note: Derived from functions, policies, CRUD and access rules; prohibitions were not authored for agents as such.

Moral aspects Derived, awaiting review

Local retirement neither deletes source data nor preserves personal data indefinitely.

Note: Sentences mentioning harm, privacy, consent or similar, collected from the specification.

Owners Filled

Steward

Accountable semantic steward role and namespace authority, identified without a commercial owner label

Roles

Semantic steward: Accountable for scope, business interpretation and proposed release decisions.
Component custodian: Maintains external metric, dimension, ontology or schema masters and approves referenced revisions.
Binding editor: Proposes local source and role bindings with provenance and declared loss.
Independent reviewer: Challenges semantic assumptions, acceptance scope and source support before promotion; absent in this run.
Access and records authority: Approves disclosure, exceptions and retention disposition in the adopting profile; executes enforcement externally.

Links to other meta-models Filled

aligned

WM-DAT-004 - Candidate taxonomy alignment to the parent Data Schema / Data Contract; bindings use pinned external constraints. No mandatory producer-consumer agreement or executable inheritance is assumed.

references

WM-DAT-001 - Dataset role and source snapshot reference; records and catalogue lifecycle remain in the dataset master.
WM-DAT-013 - Select metric revisions and input roles; formula ownership, generic aggregation semantics and metric lifecycle remain external.
WM-DAT-014 - Select dimension roles and eligible hierarchy levels; member and hierarchy maintenance remain external.
WM-KNW-001 - Pin concept and ontology references without importing ontology governance or arbitrary entailment into local identity.
WM-DAT-011 - Optional assertion graph or evidence view; the semantic definition does not own instance graph contents.
WM-DAT-005 - Optional mapping implementation and run evidence references; transformation and runtime lineage stay with their masters.

neighbor

WM-DAT-004 Data Schema / Data Contract - The registry parent is a taxonomy lead. Align the specialized semantic composition to an external schema or contract; do not infer a producer-consumer agreement from the existence of business meaning. Generic schema constraints and contractual obligations stay with that master. No executable inheritance or mandatory contract instance is ratified.
WM-DAT-001 Dataset - A semantic definition can describe several datasets or an intended dataset role. A source binding references the selected asset and revision, while records, access enforcement and asset lifecycle remain external.
WM-DAT-013 Metric Definition - This model selects a metric revision and binds its inputs and permitted dimensions. Formula, generic aggregation semantics and metric ownership remain in the metric master. A local override that changes meaning requires a separately governed metric revision.
WM-DAT-014 Dimension and Hierarchy - This model assigns analytical roles, permitted levels and context restrictions to pinned dimension references. It does not maintain member lists, hierarchy edges or their lifecycle.
WM-KNW-001 Model / Ontology and WM-DAT-011 Knowledge Graph - Term and ontology references constrain interpretation but do not turn this definition into a universal ontology or the graph of instance assertions. Graph publication is a projection, not the aggregate root.
WM-DAT-005 Data Pipeline and external evaluators - Mapping and assessment references document intended interpretation and supplied evidence. They do not own data transformation runs, query plans, metric calculation, evaluator internals or enforcement state.

parent

WM-DAT-004

What else AI and robots need to interact with it Incomplete

Sources Filled

RDF 1.1 Concepts and Abstract Syntax - W3C
SKOS Simple Knowledge Organization System Reference - W3C
The RDF Data Cube Vocabulary - W3C
Shapes Constraint Language (SHACL) - W3C
PROV-O: The PROV Ontology - W3C
R2RML: RDB to RDF Mapping Language - W3C
RFC 3339: Date and Time on the Internet: Timestamps - IETF
SDMX Standards Section 2: Information Model - UML Conceptual Design - SDMX initiative
ODRL Information Model 2.2 - W3C

Open questions

Restore independent external review before canonical or publishable-draft promotion.
Run the coordinator HTTP checker, verify exact source versions and errata, and clarify the SDMX chronology without treating HTTP success as claim verification.
Develop profile-specific mappings and acceptance fixtures for ambiguous joins, non-additive metrics, historical dimensions, missingness, empty validation targets and restricted disclosures.
Pin external metric, dimension, dataset, schema and ontology specifications and test change impact before operational adoption.
Independent second-provider review is absent; local self-audit cannot replace it.
Full source-version, errata, rights and current-profile verification remains incomplete. SDMX revision-history dates require clarification; no chronology is inferred beyond the stated cover version.
Executable nested instance schemas, pinned neighbor specifications and acceptance fixtures are not implemented.
Domain-specific aggregation, fiscal calendars, exchange rates, slowly changing dimensions and disclosure thresholds require qualified profiles.
Non-relational mappings and formal reasoning profiles need additional evidence and testing.

Machine files

Card as JSON Specification AGENTS.md Source Card backlog vercy.card/1

Provenance

world-models research · reviewable-draft

Built from: models/wm-dat-015-semantic-model/spec.yaml