Published

Concept / Term

vr.wm-knw-006 · wm-knw-006-concept-term

Represent a stable, identified semantic concept as a unit of thought that is independent of any label, language or storage format, together with the designations, definitions, notes and usage descriptions that describe it, so that agents and systems can refer to the same meaning across languages, vocabularies and interfaces.

World Models Information and virtual systems INF.KNW.CON

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

9 bundles · 19 layers · 40 findings · 173 questions

Concept identity and delimitation What makes this concept a single, citable unit of thought: how it is identified, what fixes its boundaries against neighbouring concepts, what kind of concept it is, and which real-world referents it denotes.

Identity and delimitation

Identifier assignment and the intensional and extensional criteria that fix the concept's boundary and drive split, merge and duplicate decisions.

Concept identifier and identity basis

The concept holds a persistent identifier independent of any label, definition wording, language or scheme. Records the assigning master system, an optional governed IRI, scheme-scoped notations, and the rule set that separates identity-preserving edits from identity-breaking changes.

Which authoritative master system assigns this concept's primary identifier, and what is that identifier value? identity
Does the concept have a governed IRI, and which body controls the namespace it is minted in? interoperability
Which notations or classification codes denote this concept, and in which coding scheme is each one valid? classification
Which changes preserve the identifier, and which force a new concept identifier rather than an edit? constraint

Intension, extension and delimiting characteristics

The characteristics an object must have to fall under the concept, how far the extension reaches, and which characteristic distinguishes the concept from its nearest confusable neighbour. Supplies the operational test for duplicate detection and for split and merge decisions.

Which essential characteristics must an object have for it to fall under this concept? definition
Which delimiting characteristic separates this concept from the nearest concept it is confused with? constraint
How is the extension bounded — open, enumerated or sampled by exemplars? composition
What evidence supports this delimitation — attested usage, expert ruling or a cited source clause? evidence
When two records share designations, which test decides whether they are one concept or two? decision

Concept kind and referents

Typing the record against its neighbours and recording which real-world objects it denotes, without absorbing those objects' own master data.

Concept kind and neighbour typing

Classifies the record as a general concept covering many objects or an individual concept covering exactly one, and asserts explicitly what it is not: not a lexical entry, not a word form, not a sense, not an OWL or RDFS class, not the referent, and not a free-text keyword. Prevents category errors during projection.

Is this a general concept covering many objects, or an individual concept covering exactly one? classification
When projected into a formal ontology, is this an individual of skos:Concept, an owl:Class, or both under a documented punning rule? interoperability
Which neighbouring construct — lexical entry, word form, sense, referent, class or keyword — has previously been conflated with this record? exception
What distinguishes this concept record from an uncontrolled keyword or tag carrying the same string? definition

Referents, exemplars and denotation

Which objects or entity records the concept denotes, and which exemplars illustrate it without being equated with it. Keeps the ISO 1087 object and concept split and the OntoLex reference and evokes split intact, holding referents as external references only.

Which real-world objects or entity records does this concept denote, and in which model are those referents mastered? relationship
Which exemplar, type specimen or illustration is used to show the concept without being equated with it? evidence
Does any designation denote the referent directly rather than through the concept, and how is that recorded? interoperability
What is explicitly not a referent of this concept despite being commonly assumed to be one? exception

Designations and linguistic form The signs that represent the concept — terms, appellations, proper names and symbols — with their acceptability ratings, their own identity where needed, their language and script tagging, their lexical description, and cross-language equivalence at the designation boundary.

Designation inventory and identity

Which signs represent the concept, how they are rated and typed, and when a designation must become an identified object in its own right.

Designation inventory and acceptability rating

The set of designations attached to the concept, each typed as term, appellation, proper name or non-linguistic symbol, each rated preferred, admitted or deprecated, and each marked as displayed or hidden for retrieval only. Carries the uniqueness constraint on preferred designations per language tag.

Which designation is rated preferred in each language, and which are admitted, deprecated or hidden? classification
What designation type is each sign — term, appellation, proper name or non-linguistic symbol? definition
Which rating-authority reference is carried on the designation, given that the rating process itself is governed outside this area? authority
Can two designations share the same rating in one language, and which constraint prevents duplicate preferred labels? constraint
Which non-linguistic symbol represents the concept, and in what medium is that symbol fixed? composition

Designation as an identified object

When a designation must carry its own attributes — rating history, defining source, grammatical data, relations to other designations — it is reified as an identified object rather than held as a plain string. Covers the SKOS-XL Label, the ISO 25964 preferred and non-preferred term classes and the TBX term section, and the constraint that a reified label has exactly one literal form.

Does this designation carry its own identifier, and which system assigns it? identity
Which attributes are held on the designation rather than on the concept, and why does each belong there? composition
How is a plain-string projection kept consistent with the reified designation's single literal form? validation
Which relations hold between designations of the same concept, such as abbreviation of, variant of or transliteration of? relationship

Language, script and lexical form

How every designation, definition and note is language-tagged and scripted, and how the linguistic description of a designation is recorded without leaking into the concept.

Language tag, script, region and transliteration

Every language-bearing value carries a BCP 47 tag built in the normative subtag order; the script subtag is taken from ISO 15924 and stated only when it is not suppressed for the language. Covers region and variant subtags, transliteration and romanization provenance, and language fallback behaviour.

Which complete BCP 47 language tag applies to this designation, including script, region and variant subtags where they are meaningful? requirement
Is the script subtag suppressed for this language, and if it is present anyway, what justifies including it? constraint
Which transliteration or romanization scheme produced this designation, and what is its source designation? provenance
Where a designation is region-specific, which territory or usage community does it apply to? spatial
How should a consuming system behave when no designation exists for the requested language tag? exception

Lexical entry, word forms and grammatical description

The linguistic description subordinate to a designation: the lexical entry realising it (word, multiword expression or affix), its canonical form, other forms, written and phonetic representations and grammatical categories. Attached at the designation boundary because symbols and appellations may have no lexical entry at all.

Which lexical entry realises this designation, and is it a single word, a multiword expression or an affix? classification
What is the canonical form of the designation, and which other forms are recorded alongside it? composition
Which grammatical categories are asserted, and from which data-category registry are their values drawn? interoperability
What evidence fixes the pronunciation or spoken form of this designation? evidence
Which forms must not be treated as separate designations of the concept? constraint

Cross-language equivalence

How designations in different languages are asserted to represent the same concept, and how gaps and partial equivalence are recorded.

Multilingual equivalence at the designation and sense boundary

Assertions that designations in different languages represent this one concept, with the degree of equivalence, the level at which the claim was made (concept-to-concept or sense-to-sense), the evidence behind it, and the way a missing target-language designation is recorded without inventing a term. Concept-to-concept mapping across schemes is deliberately excluded.

Which designations in other languages are asserted to represent this same concept, and at what degree of equivalence? relationship
Where a target language has no designation, how is that gap recorded without coining a term? exception
Is equivalence asserted concept-to-concept or sense-to-sense, and what follows for downstream reuse? decision
What evidence supports each equivalence assertion? evidence
Which equivalence claims are known to be partial or contested, and how is that flagged? quality

Definition, scope and usage The written descriptions that make the concept understandable and usable: definitions, typed notes, subject field and homograph disambiguation, and the usage context and attestation of its designations.

Definitions and notes

The intensional statement of meaning and the typed documentation attached around it.

Definition of the concept

The statement that describes the concept and differentiates it from related concepts, recorded per language with a definition type, the delimiting characteristic it carries, the source it is taken or adapted from, and checks against circular and purely negative definitions.

What is the definition text in each language, and which language tag applies to each version? definition
Is the definition intensional, extensional or ostensive, and which delimiting characteristic does it carry? classification
From which source document, clause or authority is the definition taken or adapted? provenance
How many definitions may coexist per language, and which rule resolves competing definitions? constraint
Which check confirms that the definition is neither circular nor a bare negation of another concept? validation

Scope notes, examples and explanatory notes

Typed documentation attached either to the concept or to a specific designation: scope note, example, general note and editorial note. Separates meaning-bearing notes that must survive interchange from editorial notes that may be filtered, and keeps change and history notes with the lifecycle and provenance areas.

Which note types are recorded, and does each attach to the concept or to a specific designation? composition
What does the scope note state as included in and excluded from the concept's coverage? definition
Which examples illustrate correct and incorrect use of the designation? evidence
Which notes are meaning-bearing and must survive interchange, and which are editorial and may be filtered? interoperability

Subject field and usage

The domain that bounds the concept and disambiguates homographs, and the audience, register, territory and attested usage of its designations.

Subject field and homograph disambiguation

The field of special knowledge in which the concept and its terms are valid, drawn from an external classification. Used to disambiguate homographic designations across domains and to bound the intension; the classification itself is referenced, never redefined here.

Which subject field does this concept belong to, and from which classification is that value drawn? classification
Which other concept shares a designation with this one in a different subject field, and how is the pair distinguished? identity
Does the subject field change the concept's delimitation, or only the usage of its designations? decision
Which classification is authoritative for subject-field values, and what happens when it is revised? interoperability

Usage context, register, audience and attestation

How and by whom a designation is actually used: audience and register, territory of currency, discouraged or offensive usage with directed alternatives, and attested occurrences that evidence real use. Keeps a designation's usage status distinct from the concept's own status, which belongs to the lifecycle area.

For which audience and register is this designation appropriate — specialist, general public, regulatory or internal? classification
Which attested occurrence evidences that the designation is really used in this sense? evidence
In which territory or community is the designation current, and where is it not used? spatial
Which usage is flagged as discouraged or offensive, and which alternative is directed instead? exception
How is a designation's usage status kept distinct from the concept's own status? constraint

Scheme placement and grouping Where a concept sits: which concept schemes assert it, whether it is a declared entry point into a hierarchy, and which non-hierarchical grouping constructs (collections, arrays, concept groups, classification levels) contain it.

Concept-scheme membership and entry points

Assertions binding a concept to concept schemes and declaring it a top concept, together with the rules governing multi-scheme participation, non-propagation of membership and hierarchy entry.

Concept-scheme membership and multi-scheme participation

SKOS places no cardinality constraint on membership: there are no conditions preventing a concept from taking part in zero, one or more than one concept scheme, skos:inScheme is not declared transitive, and a semantic link between two concepts does not entail that both are in the same scheme. Membership must therefore be asserted per concept per scheme, and an implementation that derives it from relations is making a local rule, not a standard entailment. skos:topConceptOf is a sub-property of skos:inScheme, so a top-concept declaration also asserts membership.

Which concept schemes does this concept assert membership in, and is one of them designated the maintenance scheme that issues its authoritative identifier? composition
Does this deployment permit a concept to belong to more than one scheme, and what local rule resolves conflicting structure between those schemes? constraint
Is membership asserted explicitly for every concept, or derived from broader, narrower or related links in this deployment? validation
How are concepts that assert no scheme membership treated by navigation, validation and export? exception

Top concepts and hierarchy entry points

skos:hasTopConcept has domain skos:ConceptScheme and range skos:Concept, and skos:topConceptOf is its inverse and a sub-property of skos:inScheme. Critically, the SKOS Reference defines no integrity condition preventing a declared top concept from also asserting a broader concept in the same scheme. A top concept is therefore a declared entry point chosen by the scheme, not a root derived from the absence of a broader link; treating it as a derived root is a common but unsupported assumption.

Which schemes declare this concept a top concept, and is that declaration made by the scheme rather than inferred from missing broader links? relationship
Does a concept declared as a top concept also assert a broader concept within the same scheme, and how is that reconciled? constraint
Which entry points should an agent use to enter a scheme's hierarchy when no top concept has been declared? decision
How is consistency kept when a concept is a top concept in one scheme but a mid-level concept in another? validation

Grouping, array and level constructs

Non-hierarchical grouping of concepts: SKOS collections and ordered collections, ISO 25964 thesaurus arrays and concept groups, and XKOS classification levels, all of which are formally distinct from concepts and must not be exported as retrievable concepts.

Collections, arrays, concept groups and classification levels

skos:Collection and skos:OrderedCollection group concepts through skos:member and the functional skos:memberList, with every list item also a member (S36); collections are disjoint from both skos:Concept and skos:ConceptScheme (S37). ISO 25964 adds ThesaurusArray with superOrdinate and subordinateArray for sibling grouping under a node label or facet indicator, and ConceptGroup with microThesaurusOf for domain subsets. XKOS adds ClassificationLevel with levels, depth and numberOfLevels. Because grouping nodes are not concepts, a hierarchical link between a parent concept and its children must still be asserted concept to concept; relying on the array to carry the hierarchy silently breaks navigation.

Which collections, thesaurus arrays or concept groups include this concept, and is any of those memberships ordered? composition
How is the disjointness of grouping constructs from concepts enforced so that a grouping node is never resolved or exported as a concept? constraint
Does a containing array carry a node label or facet indicator that organises siblings without itself being a retrievable concept? classification
Does this concept sit at a declared classification level, and what depth and ordering does that level carry? measurement
When a grouping construct sits between a parent concept and its child concepts, is the concept-to-concept hierarchical link still asserted directly? validation

Within-scheme semantic relations Relations asserted between concepts inside one concept scheme: direct hierarchical links and their derived transitive closure, refined hierarchical kinds, associative relations and their disjointness from hierarchy, and the modelling of compound (pre-coordinated) concepts.

Hierarchical and associative relations

The direct relation assertions a concept makes inside its scheme, the refinement of hierarchical kind, the associative relation, and the transitivity, symmetry and disjointness rules that constrain all of them.

Direct hierarchical assertion versus derived transitive closure

skos:broader and skos:narrower are inverses and are, by convention, used only to assert direct hierarchical links; neither is declared transitive. skos:broaderTransitive and skos:narrowerTransitive are the transitive properties that capture direct or indirect links. The operational consequence is that ancestor and descendant sets are derived views: they may be computed at query time or materialised for performance, but must never be written back as if they were asserted direct links, because that destroys the ability to reconstruct the authored hierarchy.

Which broader and narrower concepts does this concept assert, and is every one of them a direct link rather than an ancestor or descendant? relationship
Is the transitive form of the hierarchy computed at read time, materialised in storage, or both, and how is materialised closure marked as derived? decision
Does this concept assert more than one broader concept, and does the governing scheme permit polyhierarchy? constraint
How are cycles and self-referential links detected in the hierarchical graph before they enter the derived closure? validation
May a hierarchical link point at a concept in a different scheme, and how is such a link distinguished from a cross-scheme mapping? interoperability

Hierarchical relation kinds: generic, partitive and instantial

SKOS offers a single undifferentiated broader relation. ISO 25964, through iso-thes, refines it into broaderGeneric, broaderPartitive and broaderInstantial, and XKOS distinguishes specializes/generalizes from isPartOf/hasPart. ISO 1087 gives the corresponding terminology-science distinction between generic and partitive relations. The refinement matters operationally because chaining behaves differently: generic subsumption composes reliably, whereas partitive chains are unsafe across changes of part-whole kind. Projecting a refined kind onto plain broader for interchange is lossy and must be declared.

What kind of hierarchical relation is asserted to each broader concept - generic, partitive, instantial or undifferentiated? classification
Which test must be satisfied before a hierarchical relation is asserted, and is the justification recorded with the assertion? requirement
Which hierarchical kinds may be chained transitively in this deployment, and at which point must a chain stop? constraint
What information is lost when a refined hierarchical kind is projected onto plain broader or narrower for interchange? interoperability

Associative relations and disjointness from hierarchy

skos:related is symmetric and is not transitive, and integrity condition S27 declares it disjoint with skos:broaderTransitive - so two concepts connected by any hierarchical path must not also be linked associatively. XKOS refines association into causal (causes, causedBy), sequential (precedes and succeeds, which are transitive, versus previous and next, which are not) and temporal (before, after). Because symmetry is a property of the relation rather than of the record, a deployment must decide once whether the inverse statement is stored or inferred, and apply that decision uniformly.

Which associative relations does this concept assert, and is a refined association subtype recorded for each? relationship
How is the rule that an associative link must not coexist with a hierarchical path between the same two concepts checked? validation
Is the inverse of each associative assertion stored explicitly, inferred at read time, or both? decision
Which refined associative relations may be composed transitively and which are explicitly non-transitive? constraint

Compound concepts and decomposition

Whether a concept is pre-coordinated, how its components are recorded without over-claiming, and how one-to-many compound equivalence is expressed when a concept can only be matched by combining several concepts in another vocabulary.

Compound concepts, coordination choice and decomposition cautions

A knowledge organization system may pre-coordinate a compound concept as a single unit or leave the combination to retrieval time (post-coordination). The SKOS Primer records that established patterns for pre-coordination have not emerged in the SKOS community and recommends extensions rather than additions to the core vocabulary, so a deployment must state its own rule. OntoLex decomp offers subterm, constituent and Component, but this decomposes a lexical entry, not a concept: a recorded lexical decomposition must never be read as a claim that the concept is the semantic composition of its parts.

Is this concept a pre-coordinated compound, and which concepts are recorded as its components? composition
Under which rule was this combination modelled as one concept rather than left to be combined at retrieval time? decision
Is the recorded decomposition lexical or semantic, and what must a consumer not infer from it? definition
How are the compound and its components linked without implying a hierarchical or equivalence claim between them? relationship

Compound one-to-many equivalence across vocabularies

ISO 25964-2 recognises that equivalence between vocabularies is often not one to one and provides compound equivalence, where a concept in one vocabulary is expressed by combining several concepts in another. The iso-thes vocabulary carries this as CompoundEquivalence with plusUF and plusUse, plus SplitNonPreferredTerm for a concept that is not directly present but is representable by combining preferred terms. SKOS has no native construct for this: a consumer reading plain SKOS will see either nothing or a set of unqualified matches, so the combination operator and the degradation behaviour must both be recorded.

Which set of target concepts jointly expresses this concept, and in which vocabulary do they sit? interoperability
Which combination operator applies to the target set, and which construct carries it when the interchange format has no compound-equivalence class? requirement
How is a one-to-many equivalence degraded for consumers that accept only one-to-one matches, and is the loss flagged? exception
Is the compound equivalence reversible, and what if anything is asserted in the opposite direction? constraint

Cross-scheme mapping, alignment and interchange How a concept is connected to concepts in other schemes, aligned to formal ontology terms, and moved between systems and serializations without losing or over-stating meaning.

Cross-scheme mapping relations and mapping sets

Mapping relation types and their formal properties, the packaging of mappings into sets or correspondences with cardinality and chain rules, and the caveats that apply when mapping across languages.

Mapping relation types and their formal properties

skos:mappingRelation is a sub-property of skos:semanticRelation. exactMatch is transitive and symmetric; closeMatch is symmetric but explicitly not transitive; broadMatch and narrowMatch are sub-properties of skos:broader and skos:narrower as well as of mappingRelation, so they participate in the hierarchical closure; relatedMatch is a sub-property of skos:related and is symmetric. Integrity condition S46 makes exactMatch disjoint with broadMatch and relatedMatch. Two facts are commonly missed: cross-scheme use of the mapping properties is a convention, not a formal requirement, and SKOS supplies no integrity condition making a same-scheme mapping inconsistent.

Which mapping relation type is asserted towards each target concept, and which scheme owns that target? relationship
What threshold distinguishes an exact match from a close match in this deployment, and who applies it? decision
How does the deployment handle the fact that broad and narrow matches are also hierarchical properties and therefore enter the transitive hierarchical closure? constraint
How is the disjointness of an exact match with a broad match or a related match between the same pair of concepts enforced? validation
Is a mapping relation permitted between two concepts of the same scheme, and what is being asserted when it occurs? exception

Mapping sets, cardinality, binding and chain limits

Mappings are managed as sets rather than as isolated statements. XKOS provides Correspondence, which compares two schemes, and ConceptAssociation with sourceConcept and targetConcept, allowing many-to-many associations to be reified and described. ISO 25964-2 similarly organises equivalence, hierarchical and associative mappings between vocabularies. Chain composition must be conservative: exactMatch is transitive so exact chains compose, but closeMatch is not transitive and a chain mixing types must not be composed at all. Every set binds to specific releases of the two schemes it compares; the releases themselves are issued by the lifecycle area.

Which mapping set or correspondence contains this concept's mapping assertions, and which two schemes does that set compare? composition
What cardinality does the mapping set permit, and how are many-to-many associations represented? classification
Which specific scheme releases does the mapping set bind to on each side? interoperability
Which mapping types may be composed across two sets to derive a third mapping, and where must composition stop? constraint
How are contradictory mappings from different sets for the same concept pair detected and reported? quality

Multilingual and cross-lingual mapping caveats

A concept is a language-independent node: SKOS attaches labels in many languages to one concept and permits only one preferred label per language tag, so a translation is normally a label of the same concept rather than a mapping to a different concept. Where two vocabularies are maintained in different languages, ISO 25964 treats cross-lingual equivalence as graded - exact, inexact, partial, one-to-many or absent. BCP 47 tags identify the language, script and region of a string; matching tags on two labels says nothing about whether the labels denote the same concept, and equal tags must never be used as evidence of equivalence.

Is this a single language-independent concept carrying multilingual labels, or are language-specific concepts linked by cross-lingual mapping? definition
What degree of cross-lingual equivalence is recorded when the target language has no exact equivalent? classification
Which BCP 47 language tags, including any script and region subtags, scope each side of a cross-lingual assertion? identity
What must a consumer not infer from two labels carrying the same language tag or from a translated label? constraint

Formal alignment and interchange

Connecting the concept to formal ontology terms with an explicit statement of entailment, and moving the concept's relation and mapping structure between systems and serializations with declared closure, crosswalks and losses.

Semantic alignment to ontology classes, properties and individuals

Aligning a concept to an ontology term is not the same as mapping it to another concept. OWL 2 gives owl:equivalentClass a set-extensional meaning over individuals, rdfs:subClassOf transitive and reflexive subsumption, and owl:sameAs identity of individuals, and each licenses a reasoner to draw consequences the vocabulary editor may not intend. OntoLex sits deliberately between the two worlds: LexicalConcept is a subclass of skos:Concept and is connected to entries by evokes and lexicalizedSense, while denotes and reference point at ontology predicates. DCAT shows the practical consequence, allowing a taxonomy to be a concept scheme, a collection or an ontology. The alignment predicate and its entailment scope must therefore be recorded with the alignment, not assumed.

Is the alignment target an ontology class, a property, an individual or another concept? classification
Which alignment predicate is asserted, and what entailment does it commit a consuming reasoner to? constraint
Why is this concept not asserted as identical or class-equivalent to the ontology term by default? definition
How is the concept connected to lexical entries and senses without importing lexical structure into the concept record? relationship

Import, export, projection and crosswalk behaviour

RDF 1.1 establishes that the abstract structure of identifiers, values and statements is independent of any concrete serialization, and that identifiers have global scope, so JSON, YAML, Markdown, tabular files, Turtle and JSON-LD are projections of one assertion set. Interchange therefore turns on three decisions that must be declared rather than implied: which closure is included (concept alone, direct relations, whole scheme, or transitive neighbourhood); how references that leave the closure are represented so they are not read as missing data; and which constructs are lossy in the target model - refined hierarchical kinds, compound equivalence and reified associations are the usual casualties. Namespace and identifier resolution follow the published-vocabulary recipes: term identifiers stay stable while version information moves into the served document.

What closure does an export include - the concept alone, its direct relations, its whole scheme, or its transitive neighbourhood? composition
How are references to concepts outside the exported closure represented so they are not mistaken for missing or broken data? exception
Which crosswalk rule maps each local relation, grouping and mapping construct onto the target interchange model? interoperability
Which constructs are lossy on a round trip through this projection, and how is the loss declared to the consumer? validation
Which namespace and identifier-resolution rule applies to concept references inside the exchanged package? identity

Concept lifecycle, decision authority and change control How a concept record enters the register, moves between governed states under a named authority, is bound to releases, and accumulates an immutable change history including merge, split and redirect.

Registration states, authority and retirement

The governed state set, who may move a record between states, and how a concept leaves active use without its identifier ever being reused.

Registration state model and transition rules

The enumerated states a concept record may occupy, split into lifecycle states (development and usage preference) and documentation states (no further development), with entry criteria, permitted transitions and the meaning each state guarantees to consumers.

Which registration states may this concept record occupy, and which are lifecycle states rather than documentation-only states? state
What mandatory attributes and associations must be complete before the record may progress beyond the recorded state? requirement
Which role must authorise each transition, and which transitions are irreversible? authority
What stability guarantee does each state give a downstream consumer about the concept's meaning? classification

Decision authority and proposal workflow

Who holds decision authority over this concept, the proposal-review-approval steps a change must pass, and the auditable record of each governed decision including dissent and appeal.

Who holds decision authority over this concept, and under what documented mandate? authority
What proposal, review and approval steps must a change pass before publication? process
What evidence, options considered and dissent must a decision capture to remain auditable? decision
How is a contested or appealed decision recorded without rewriting the original decision? exception

Deprecation, retirement and succession

Grounds for deprecating rather than correcting a concept, the successor relationship offered to clients, the guarantee that a retired identifier is never reused or repointed, and how long resolution is promised.

On what grounds may this concept be deprecated instead of corrected in place? lifecycle
Which successor concept replaces the deprecated concept, and is the replacement authoritative or merely advisory? relationship
What must a client still holding the deprecated identifier do, and until when is resolution guaranteed? interoperability
How is a retired identifier prevented from ever being reused or repointed to a different meaning? identity

Versioning, change history and duplicate resolution

How a concept record's state is bound to an immutable release, how every change is recorded append-only with cause and time, and how duplicates are detected and resolved by merge, split or redirect.

Version designation and release binding

Which immutable release of the owning scheme a governed record state belongs to, how a change is classified as editorial, additive or breaking, and how a version designation stays distinct from record identity.

Which immutable release of the owning concept scheme does this governed record state belong to? identity
How is a proposed change classified as editorial, additive or breaking, and who performs that classification? classification
What guarantees that a published release remains unaltered and retrievable at a later date? constraint
How does a version designation relate to, and remain distinct from, the concept record identifier? definition

Change history, event timing and audit trail

Append-only atomic change records with the causing activity, agent and decision, separated event, effective and ingestion times, and the ability to reconstruct the record as it stood at any past instant.

What atomic change was made, to which element of the concept record, and from what prior value? event
Which activity, agent and prior decision caused this change? provenance
When did the change occur, when does it take effect, and when was it ingested into this register? temporal
How can a consumer reconstruct this concept record exactly as it stood at an arbitrary past instant? evidence
How is an erroneous change record corrected without deleting or rewriting history? quality

Duplicate candidacy, merge, split and redirect

Evidence by which another record is judged to denote the same concept, which identifier survives a merge, how a split disambiguates inbound references, and how automated duplicate signals are bounded.

By what evidence is another record judged to denote the same unit of thought as this one? evidence
Which identifier survives a merge, and how are the non-surviving identifiers redirected? decision
When a concept is split, how are prior usages and inbound references disambiguated across the successors? composition
Which automated duplicate-candidate signals are used, and what false-positive tolerance is accepted before human review? measurement

Provenance, evidence, validation and editorial quality Where the concept's meaning comes from, what warrant justifies it, which constraint sets it is checked against, and how editorially complete and mature the record is.

Authoritative sources and evidentiary warrant

The external authority for the concept's definition, the derivation relationship to it, and the warrant that justifies admitting the concept at all.

Authoritative source and derivation provenance

Which external source is authoritative for this concept's definition and at which version, whether the concept is originated, adopted verbatim, adapted or derived, and how later change or withdrawal of that source is reflected.

Which external source is authoritative for this concept's definition, and at what version or publication date? provenance
Is this concept originated locally, adopted verbatim, adapted, or derived from another vocabulary? classification
Which agent asserted the current definition, and on whose behalf did that agent act? ownership
How is a later change to, or withdrawal of, the authoritative source reflected in this record? lifecycle

Warrant and rationale evidence

The governance-grade evidence set justifying that this is a distinct unit of thought with the scope claimed, held by reference to retained attestations and citations rather than restated locally.

What warrant justifies admitting this as a distinct concept rather than a variant of an existing one? evidence
Which retained attestations and citations support the current definition, and where are they held? composition
What counter-evidence or rejected alternative scoping was considered, and why was it set aside? decision
When supporting evidence cannot be published, what minimal defensible reference is exposed instead? privacy

Validation, conformance and editorial quality

Which constraint sets apply and what the last check returned, and how editorially complete and mature the record is on a governed scale.

Constraint declaration and conformance reporting

Which constraint sets are declared applicable at which severity, the outcome of the most recent check against a named record state and constraint version, and what conformance is claimed against external standards.

Which constraint sets apply to this concept record, at which severity, and who declared them applicable? validation
What did the most recent check return, against which record state and which constraint-set version? quality
Which failures block publication and which are advisory only? constraint
What conformance to an external standard is claimed for this record, and on what evidence? interoperability

Editorial completeness, maturity and quality measurement

Which editorial criteria the record has met, how mature and stable the definition is on a governed scale, which quality dimension each measurement addresses, and when the next editorial review falls due.

Which editorial completeness criteria has this record met, and which remain open? measurement
How mature and stable is the definition, expressed on a governed scale? classification
Which quality dimension does each recorded measurement address, and by which metric was it computed? quality
Who last reviewed the editorial content, and when is the next review due? temporal

Stewardship, namespace, rights and retention Who owns and maintains the concept record, who controls the namespace its identifier lives in, and under what rights, access, embargo and retention terms the record and its evidence are held.

Ownership, stewardship and namespace authority

The owning organisation, the named steward, delegated responsibilities and escalation, and the namespace that mints and guarantees the identifier.

Ownership, stewardship and delegation

Which organisation owns the record, which named steward maintains it, what is delegated without transferring ownership, how escalation works when a steward is unavailable, and how an ownership transfer is recorded without breaking identifier continuity.

Which organisation owns this concept record, and which named steward maintains it? ownership
What is the escalation and hand-over path when the steward is unavailable or the role lapses? process
Which responsibilities are delegated to a contributor without transferring ownership? authority
How is a change of owning organisation recorded without breaking identifier continuity? lifecycle

Identifier and namespace stewardship

Which namespace mints the concept identifier, who controls it, what resolution and persistence is committed, how identifiers are minted and prevented from collision or reuse, and what happens if the namespace is transferred or discontinued.

Which namespace mints this concept's identifier, and which body controls that namespace? identity
What resolution and persistence commitment is made for the identifier, and for how long? constraint
How are identifiers minted, reserved and prevented from collision or reuse across the namespace? process
What happens to identifiers if the namespace is transferred, forked or discontinued? exception

Rights, access, embargo, retention and disposition

Under what licence and access terms the record and its evidence may be reused, what is embargoed until when, and which retention class and disposition instruction the record carries for execution elsewhere.

Licence, access rights and embargo

The licence governing reuse of the concept record and its attached evidence, which parts are restricted and to whom, what is embargoed and what releases it, and whether attached evidence identifies a living person or carries third-party rights.

Under which licence may this concept record and its attached evidence be reused? access
Which parts of the record are restricted, to which parties, and on what stated basis? security
Is any content embargoed, and which event or instant releases the embargo? temporal
Does any attached evidence identify a living person or carry third-party rights that constrain publication? privacy

Retention class, disposition instruction and tombstone

The retention class and disposition instruction the concept record carries, legal-hold state, the tombstone that must survive any disposition, and which authority executes the disposition.

Which retention class applies to this record and to each attached evidence item? retention
Which authority approves and executes disposition, and what evidence of execution is returned? authority
Is the record under a hold that suspends disposition, and what releases that hold? exception
What minimum tombstone must survive after disposition so that references do not silently break? requirement

Classifiers Filled

Family: World Models
Category: Information and virtual systems
Entry kind: entity
Navigation path: NAV.INF.KNW.CON
Domain: INF.KNW.CON
Industry: Cross-industry
Tags: conceptterminf.knw.con

What it is Filled

WM-KNW-006 models the concept as an aggregate root with its own persistent identity, plus the subordinate or referenced descriptions that give it meaning: designations (terms, appellations, proper names, symbols) with acceptability ratings, lexical entries and word forms, definitions and typed notes, subject field, referents and usage context; the concept's placement in concept schemes and its hierarchical, associative and cross-scheme mapping relations; and its lifecycle, provenance, ownership, editorial governance, access policy references and quality assertions. The model is format-neutral: RDF/SKOS, TBX, OntoLex, JSON, YAML, Markdown, Git, MCP and document stores are projections of the same semantics. External standards are treated as alignments, not as conformance claims.

In scope

Concept identity: authoritative identifier, governed IRI, notation, and the rules that decide when a change breaks identity
Concept delimitation: intension, extension, essential and delimiting characteristics, split/merge and duplicate tests
Concept kind: general versus individual concept, and explicit separation from lexical entry, word form, sense, class/type, referent and free-text keyword
Designations attached to the concept: term, appellation, proper name and non-linguistic symbol, with preferred/admitted/deprecated acceptability ratings and hidden retrieval-only labels
Designation-level identity and attributes where terms are reified rather than carried as plain strings
Language, script, region and variant tagging of every designation, definition and note, plus transliteration and romanization
Lexical description at the designation boundary: lexical entry, canonical form, other forms, written and phonetic representation, grammatical categories
Definitions and typed notes: scope note, example, general note, editorial note, and their attachment level
Subject field assignment and homograph disambiguation across domains
Referents and exemplars denoted by the concept, held as references to external entity records
Usage context: audience, register, territory, discouraged usage and attested occurrences
Multilingual equivalence asserted at the designation and sense boundary
Concept scheme membership, top-concept status and scheme-scoped constraints
Hierarchical, associative and instantial semantic relations between concepts
Cross-scheme concept mappings and their degree of match
Concept and designation lifecycle: status, supersession, deprecation, redirect and versioning
Provenance, ownership, editorial authority and change history
Access-policy references, quality assertions and validation profiles for concept records

Out of scope

The real-world objects and entity records that a concept denotes; their attributes, identity resolution and lifecycle belong to domain entity models
Formal ontology axiomatisation and reasoning: OWL/RDFS class semantics, restrictions, entailment and consistency checking
Publication of a full lexical or lexicographic resource (dictionary entries, sense inventories, corpora, annotation layers) beyond the designation-level description carried here
Documents, records and content items that merely mention or are indexed with a concept
Search, retrieval, ranking and indexing engines that consume designations
Translation-management workflow, translation memory segments and localisation project state
Identifier minting and resolution infrastructure such as PID, handle or DOI services
Runtime access-control evaluation, policy enforcement and audit-trail storage; this model carries policy and audit-record references only
Machine-learning embeddings, vector representations and similarity models derived from concepts
Retention scheduling, legal hold and physical destruction execution, which belong to the adopting Dimension's records policy

Why it exists Filled

Distinguishing features Filled

A concept is a unit of meaning independent of any label, language or storage format.
Unlike a lexical entry, it does not model word forms or senses of a language.
Unlike an ontology class, it carries no formal axioms for reasoning.
Unlike a free tag, it has stable identity, definitions and governance.

What robots and AI may and may not do Filled

Must not

Change a concept's meaning while keeping its identifier.
Edit a published release in place.
Advance a lifecycle state without recorded authority.
Publish evidence that identifies a living person without permission.
Claim conformance to an external vocabulary standard without evidence.

Only with a human decision

Approving a new or changed definition.
Deprecating a concept that others depend on.

May

Establish a concept identity and attach designations in several languages.
Record definitions and typed notes.
Resolve a label to candidate concepts with ambiguity shown.
Assert relations within a scheme and mappings across schemes.

Moral aspects Filled

Terms used for people and groups can stigmatise; labels should be reviewed for respectful usage.
Biased definitions propagate into systems that classify people.
Minority languages deserve equal standing in designations.

Who is affected

Communities described by the terms
Terminologists and stewards
Users of systems built on the vocabulary

Owners Filled

Steward

The adopting package MUST declare itself owner of vr.wm-knw-006 and name a registration authority, at least one concept steward, a namespace controller and a rights-and-privacy reviewer before any concept record may progress beyond the recorded state.

Roles

Registration authority: Approve progression of a concept record beyond the recorded state and confer registration status; Issue statements of authorisation and hold the mandate recorded on decision records; Approve retirement, merge, split and redirect, and maintain the reuse-exclusion register
Concept steward: Sponsor a concept record for progression and maintain its editorial content and evidence references; Classify proposed changes as editorial, additive or breaking and refer breaking changes for decision; Respond to duplicate candidacy referrals and keep successor pointers accurate
Submitter / proposer: Open a governance case with a proposed concept or change, its warrant type and evidence references; Attach the duplicate-candidacy screening result before submission; Respond to reviewer findings without editing published record states directly
Editorial reviewer: Assess editorial completeness, definition maturity and quality measurements against the governed scales; Record findings, dissent and rejected alternative scopings for the decision record; Set and monitor the next-review-due instant
Validation operator: Declare which constraint-set versions apply and submit record states for external evaluation; Retain the returned validation report and link it to the exact validated state digest; Report blocking failures and record any accepted waiver with its authorising decision
Rights and privacy reviewer: Review evidence flagged as identifying a living person or carrying third-party rights before publication; Direct minimisation or redaction and approve the restricted-reference stub where disclosure is refused; Record the licence, access rights and embargo terms applied to each record part
Namespace controller: Mint, reserve and exclude identifiers in the model's namespace and prevent collision or reuse; Maintain the persistence and resolution commitment and publish its end conditions; Execute namespace succession if the namespace is transferred, forked or discontinued
Auditor: Reconstruct concept record states at a stated instant and verify digests, algorithms and mode versions; Test that governance records are append-only and that timestamps distinguish event, effective and ingestion time; Report integrity findings such as change-record gaps, digest mismatches or reused sequences

Links to other meta-models Filled

child

WM-KNW-002 — concept scheme or knowledge organization system (registered parent) - A concept is registered as a member of one or more schemes. This model carries only the inScheme reference and any scheme-scoped notation; scheme membership rules, top-concept designation, scheme versioning and scheme-wide editorial policy remain with the parent.
WM-KNW-002 concept scheme / knowledge-organization model (registered parent) - This concept is placed in schemes governed by the parent model. The parent owns scheme identity, maintenance agency, coverage and release policy; this model carries only the membership assertion, the top-concept declaration and the scheme reference used to bind a mapping.
WM-KNW-002 (parent knowledge organization system / concept scheme model) - Concepts governed here are registered within, and their record states bound to, releases of a scheme owned by the parent. Scheme membership, top concepts, scheme boundaries and scheme-level release identity remain with the parent; this model carries the scheme and release references only.

aligned

W3C SKOS and SKOS-XL vocabulary - Express concept identity, labels, notation and typed notes in RDF, including reified labels where designation-level attributes are needed. Alignment only: conformance is claimed solely where a validated profile and evidence exist.
W3C OntoLex-Lemon lexicon model - Attach lexical entries, canonical and other forms, written and phonetic representations and senses at the designation boundary. The lexicon's own entry inventory, sense inventory and editorial lifecycle remain in the lexical resource.
ISO 25964 SKOS extension (iso-thes) thesaurus profile - Map reified preferred and non-preferred terms and the term-level status property. The extension's hierarchical relation properties are used by the relations area of this model, not by the semantic core.
TBX interchange profile (ISO 30042 dialects) - Support concept-oriented terminological interchange of designations, grammatical data and usage notes, and drive the loss report when a dialect cannot carry an attribute. Dialect validation artefacts remain with the TBX specification.
OMG Multiple Vocabulary Facility (MVF) - Map a concept entry and its designation sets onto model elements expressed in multiple natural languages and terminologies, using MVF's normative mappings to ISO 1087 and SKOS. MVF's model-management semantics remain with MVF.
W3C SKOS and SKOS-XL vocabulary (http://www.w3.org/2004/02/skos/core#) - Relation, mapping and grouping constructs are aligned to SKOS classes and properties, and SKOS integrity conditions are evaluated as alignments. The alignment is a recorded correspondence, not a conformance claim; no SKOS conformance is asserted without evidence.
ISO 25964 thesaurus data model and the iso-thes RDF vocabulary (http://purl.org/iso25964/skos-thes) - Refined hierarchical kinds, thesaurus arrays, concept groups, micro-thesauri and compound equivalence are aligned to ISO 25964 constructs where the governing scheme uses them. Where the clause-level normative text was not verified, the alignment is recorded as provisional.
XKOS extension for statistical classifications (DDI Alliance) - Classification levels and depth, refined semantic relations, and the Correspondence and ConceptAssociation structures used for many-to-many cross-scheme and cross-version mapping are aligned to XKOS for statistical classification schemes.
Provenance activity record model (PROV-conformant Entity, Activity, Agent and derivation records) - Express governance events so they project to activity, generation, derivation and attribution structures without redefining provenance semantics or owning any provenance store.
Vocabulary status and asset version-chain vocabulary (status codes, previous/next/current version, version notes) - Map local registration status and version-chain pointers to published status and versioning terms rather than redefining them, while recording that the source note has been retired and migrated to another maintainer.
Metadata registry registration model (registration authority, submitter, steward, registration status categories) - Align local registration state names, entry criteria and progression authority to registry practice where the adopting Dimension operates a metadata registry; registry-wide administration remains there.
Data quality measurement model (quality dimensions, categories, metrics and measurements) - Express editorial quality as dimension-scoped measurements against named metrics. Metric definitions, computation and any certification remain in the quality model.

references

IANA Language Subtag Registry (IETF BCP 47) - Supplies the value space and validity rules for language, extlang, region and variant subtags, including Deprecated, Preferred-Value and Suppress-Script handling. Registry maintenance, subtag deprecation and republication remain with IANA and the IETF.
ISO 15924 script code registry (Unicode Consortium Registration Authority) - Supplies four-letter script code values for the script subtag. Code allocation, change requests and distribution remain with the Registration Authority.
Subject field or domain classification model - Supplies subject-field values used to bound the concept and disambiguate homographs. Classification structure, maintenance and revision handling remain in that model; this model stores the referenced value and its classification version.
Referent or real-world entity model - Identifies the objects the concept denotes. Entity attributes, identity resolution, matching and lifecycle remain in the entity model; this model records the denotation link and the exclusion statements.
Ontology term model (RDFS and OWL 2 classes, properties and individuals) and its reasoning environment - Semantic alignments point at ontology terms. This model carries the target reference, the alignment predicate and its declared entailment scope only. Entailment computation, reasoner configuration, closure materialisation and inference results are owned by the referenced reasoning environment.
OntoLex-Lemon lexical entry, form and sense model - Lexicalisation of the concept and lexical decomposition of multiword expressions are reached through the evokes and lexicalized-sense bridge. Entries, forms, senses, translations and label relations remain owned by the lexical model and are not reproduced here.
IANA Language Subtag Registry (BCP 47 / RFC 5646) - Language, script and region subtags that scope a cross-lingual assertion are referenced from the registry. Subtag validity, registration and registry maintenance are external; this model only records the tag and the canonicalisation rule applied before comparison.
DCAT catalogue and theme taxonomy model - A scheme may be referenced by a catalogue as a theme taxonomy and a concept used as a theme, which is why concept identifiers must be dereferenceable IRIs. Catalogue records, distributions and dataset description remain outside this model.
Party / Agent model (person, organisation or software agent acting as submitter, steward, reviewer or registration authority) - Carry an agent reference plus the role in which that agent acted for a governance event. Agent identity, contact data and organisational lifecycle stay in the party model; this model never reproduces them.
Constraint / shape specification model (versioned constraint sets and severities) - Record which constraint-set version was declared applicable and retain the returned report. Constraint language, evaluation algorithm and engine execution are owned there; a reference to the evaluator grants no ownership of evaluation.
Rights and licence policy model (machine-readable permissions, prohibitions, duties and constraints) - Carry the licence or policy reference and subject-specific constraint parameters such as embargo end. Policy interpretation, evaluation and enforcement belong to the policy model and its evaluator.
Records retention and disposition schedule model (retention classes, holds, disposition authority) - Carry the retention class reference, hold state and disposition instruction for this model's own records. Scheduling, approval and destruction execution remain with the records authority.
Persistent identifier and namespace registry model (minting, reservation, resolution and exclusion) - Carry the namespace, identifier scheme, persistence commitment and reuse-exclusion entries for this concept. Minting service operation and resolver infrastructure are owned there.

composes

WM-KNW-006 lexical designation structures (designations, language-tagged labels, definitions and ct-core-usage-attestation) - Governance states, decisions and change records attach to the stable concept while designation-level content and retained usage attestations remain in the lexical structures. This pass carries designation references, disclosure state and the publication access exception only.
WM-KNW-006 semantic relation and mapping structures (ct-rel-art-mapping-set, ct-rel-art-crosswalk-specification, ct-rel-art-interchange-package) - Mapping and relation assertions are defined and composed there; this pass supplies only the change classification, decision, validation and release-binding gates that a mapping change must pass through.

neighbor

Lexical entry and word form (OntoLex-Lemon lexicon) - A lexical entry is a unit of the lexicon with grammatically related forms and base meanings; a concept is a unit of thought. This model attaches lexical entries and forms at the designation boundary and references the lexicon, but does not own the lexicon's entry inventory or its editorial lifecycle.
Class or type in a formal ontology - SKOS concepts are individuals of skos:Concept, not classes; the SKOS Primer records that OWL DL forbids treating them as classes. A concept record may be aligned with an owl:Class under a documented punning rule, but class axioms, restrictions and entailment stay in the ontology model.
Referent or real-world object - ISO 1087 separates the object (anything perceivable or conceivable) from the concept (a mental unit combining characteristics of objects). This model records which referents a concept denotes as references; the referent's own master data belongs elsewhere.
Concept scheme, thesaurus or controlled vocabulary (parent WM-KNW-002) - The scheme owns membership, top-concept designation, scheme versioning and scheme-wide editorial policy. A concept carries an inScheme reference; the same concept may participate in more than one scheme without the scheme's rules migrating into the concept record.
Free-text keyword or tag - A keyword is an uncontrolled string with no identity, no rating, no language governance and no definition. A designation in this model is always bound to an identified concept and carries a language tag and an acceptability rating.
Lexical sense (ontolex:LexicalSense) - A sense pairs one lexical entry with one reference; equivalence asserted sense-to-sense is narrower than equivalence asserted concept-to-concept. This model records which level an equivalence claim was made at, but does not maintain a sense inventory.
Notation, code list and classification registry - skos:notation is a lexical code unique within a concept scheme, typed by datatype. The code list itself, its allocation rules and its release cycle are owned by the issuing registry, not by the concept record that cites the code.
Semantic relations, mappings, lifecycle, provenance, access and quality areas of WM-KNW-006 - These are in scope for the combined model but are specified in separate areas. The semantic core carries only the identity-breaking trigger that forces a split or merge, reference-only pointers to rating authority and definition source, and no relation, status, approval or audit machinery of its own.

parent

WM-KNW-002

What else AI and robots need to interact with it Filled

Capabilities and actions required Filled

Establish a concept identity: Create a new concept record with a persistent identifier and an initial delimitation, after confirming that no existing concept already covers the same intension.
Attach a designation to a concept: Attach a term, appellation, proper name or symbol to an existing concept with its language tag, designation type and acceptability rating, reifying the designation where the active profile requires designation-level attributes.
Record a definition or typed note: Attach a language-tagged definition or a typed note to a concept or to a specific designation, with its source reference and attachment level.
Resolve a designation to candidate concepts: Look up a string in a given language and optional subject field and return the concepts whose designations match, with the evidence needed to disambiguate homographs.
Validate a concept record against declared constraints: Check a concept record and its designations, definitions and notes against the structural constraints declared for the active profile, and return findings.
Project a concept set to or from an interchange profile: Render concepts and their designations into a named interchange profile, or read them back, producing an explicit report of anything the target profile cannot carry.
Assert a within-scheme semantic relation: Record a direct hierarchical or associative relation between two concepts inside one scheme, with its refined kind where the scheme supports refinement, and with the justification note that accompanies the assertion.
Assert a cross-scheme mapping: Record a mapping from this concept to a concept in another scheme, with the mapping type, the target scheme reference, the bound scheme releases and the basis for the chosen strength.
Derive transitive hierarchy closure: Compute ancestor and descendant sets from direct hierarchical assertions, honouring per-kind chaining rules and stopping where a refined kind changes or a scheme boundary is crossed.
Navigate a scheme from its entry points: Return a scheme's declared top concepts, or a computed entry set where none are declared, and traverse from there through direct relations and grouping constructs to a bounded depth.
Check relation and mapping integrity: Evaluate this model's own relation, mapping, membership and grouping assertions against the declared integrity conditions and produce a findings report.
Compose a mapping chain conservatively: Derive an indirect mapping between two concepts by composing mappings across sets, only where the mapping types involved licence composition.
Export a relation and mapping package: Produce an interchange package containing the selected closure of relations, memberships, mappings and alignments, projected through a named crosswalk into a target model and serialization.
Import a relation and mapping package: Ingest an external package of relation, mapping and alignment assertions, verify its integrity, resolve or retain its references, and merge accepted assertions under the local rules.
Submit concept proposal: Register a candidate concept or a proposed change to an existing concept, opening a governance case without asserting any published state.
Transition registration state: Move a concept record from its current registration state to a permitted next state under the required authority.
Record governance decision: Emit an immutable decision record capturing the question, options, outcome, participants and dissent for one governed decision.
Bind concept state to a release: Bind the current concept record state to an immutable scheme release and classify the change since the previous bound state.
Validate concept record: Submit a concept record state to an externally owned evaluation against declared constraint-set versions and retain the returned report.
Assess duplicate candidacy: Score and review whether another record denotes the same unit of thought, producing a reviewed sameness judgement rather than an automatic merge.
Execute merge or split: Apply an approved merge or split, establishing survivorship, successor scopes and redirects for every affected identifier.
Deprecate concept with successor: Mark a concept deprecated or retired, publish the obsolescence reason, successor or advisory alternatives, and the resolution guarantee.
Reconstruct concept record at an instant: Rebuild the concept record exactly as it stood at a given instant by replaying change records, for audit and dispute resolution.
Apply retention and disposition instruction: Evaluate the retention class and hold state for a record part and hand the disposition instruction to the authority that executes it.

Sources Filled

SKOS Simple Knowledge Organization System Reference - World Wide Web Consortium (W3C)
SKOS Simple Knowledge Organization System eXtension for Labels (SKOS-XL), section of the SKOS Reference - World Wide Web Consortium (W3C)
SKOS Simple Knowledge Organization System Primer - World Wide Web Consortium (W3C)
Lexicon Model for Ontologies: Community Report - W3C Ontology-Lexica Community Group
MVF ISO 1087 Vocabulary for Terms and Definitions (RDF) - Object Management Group (OMG)
Multiple Vocabulary Facility (MVF) specification page - Object Management Group (OMG)
ISO 25964 — the international standard for thesauri and interoperability with other vocabularies - National Information Standards Organization (NISO)
SKOS-Thes: ISO 25964 SKOS extension namespace documentation - Dublin Core Metadata Initiative (hosting the ISO 25964 SKOS extension by A. Isaac and J. De Smedt)
RFC 5646: Tags for Identifying Languages (BCP 47) - Internet Engineering Task Force (IETF)
ISO 15924 Registration Authority — Codes for the representation of names of scripts - Unicode Consortium (ISO 15924 Registration Authority)
RFC 3339: Date and Time on the Internet: Timestamps - Internet Engineering Task Force (IETF)
Introduction to TermBase eXchange (TBX) - TBX Info (complement to ISO 30042, maintained with LTAC Global)
ISO 1087:2019 Terminology work and terminology science — Vocabulary - International Organization for Standardization (ISO)
XKOS: An SKOS extension for representing statistical classifications - DDI Alliance
OWL 2 Web Ontology Language Primer (Second Edition) - World Wide Web Consortium (W3C)
RDF 1.1 Concepts and Abstract Syntax - World Wide Web Consortium (W3C)
Best Practice Recipes for Publishing RDF Vocabularies - World Wide Web Consortium (W3C)
ANSI/NISO Z39.19-2005 (R2010) Guidelines for the Construction, Format, and Management of Monolingual Controlled Vocabularies - National Information Standards Organization (NISO)
Data Catalog Vocabulary (DCAT) - Version 3 - World Wide Web Consortium (W3C)
PROV-O: The PROV Ontology - W3C
Data on the Web Best Practices - W3C
Shapes Constraint Language (SHACL) - W3C
Data Quality Vocabulary (DQV) - W3C
DCMI Metadata Terms - Dublin Core Metadata Initiative (DCMI)
RFC 8785: JSON Canonicalization Scheme (JCS) - IETF
ISO/IEC 11179-6:2023 Information technology — Metadata registries (MDR) — Part 6: Registration - ISO/IEC
OBO Foundry Principle 4: Versioning - OBO Foundry
Obsoleting an Existing Ontology Term (OBO Academy / OBOOK) - OBO Foundry
Asset Description Metadata Schema (ADMS) - W3C
ODRL Information Model 2.2 - W3C
Semantic Versioning 2.0.0 - Semantic Versioning (semver.org)

Open questions

Obtain licensed access to ISO 1087:2019, ISO 25964-1 and -2, ISO 30042:2019 (TBX), ISO 24613 (LMF) and ANSI/NISO Z39.19, read the clause text, and re-derive the support for the registration state model, the generic/partitive/instantial hierarchy kinds and compound equivalence rather than relying on surrogate renderings.
Retry DatCatInfo or an equivalent data-category repository and pin persistent identifiers for part of speech, term type and administrative status, which ct-core-q-lex-grammar currently leaves as an unnamed registry reference.
Research SSSOM and comparable mapping-exchange formats and either align ct-rel-art-mapping-set to them or record the non-alignment as a permanent open item; add confidence, similarity and probabilistic justification, which the model currently reduces to a free-text strength basis.
Determine, per jurisdiction, whether a designation record naming a living person constitutes personal data, and reconcile any erasure obligation with the mandatory surviving tombstone and the never-reuse identifier rule.
Resolve where the mandatory interchange loss report is held: whether it is a retained governed artifact in the ct-core zone, a section of ct-rel-art-interchange-package, or an explicitly transient response, and record the decision before identifiers are frozen.
Settle system-of-record ownership for the topConceptOf / hasTopConcept inverse pair and for inScheme membership between WM-KNW-006 and parent WM-KNW-002, and register the outcome as a relationship-contract entry so the composition ledger stops referencing entries that do not exist.
Model multilingual governance divergence, where different language communities steward the same concept under different authorities, and cover sign-language, tactile and other non-written, non-audio designation modalities that the current designation-type and symbol artifact only partially reach.
Clause text of ISO 1087:2019, ISO 704, ISO 12620-1:2022, ISO 30042:2019 and ISO 24613 (LMF) is paywalled and was not read. ISO 1087 notions are grounded through the publicly retrievable OMG rendering (SRC-005) and no ISO clause is quoted.
The DatCatInfo data category repository was unreachable during research (connection refused), so no data-category persistent identifiers for part of speech, term type or administrative status are cited; grammatical category values are declared as registry-referenced without naming specific identifiers.
Sign-language designations, tactile designations and other non-written, non-audio modalities are only partially covered by the designation type and symbol artifact; no primary source consulted specifies their representation.
No treatment of derived representations such as embeddings, ranked synonym expansion or automatically induced senses.
The clause-level normative text of ISO 25964-1 and ISO 25964-2 and of ANSI/NISO Z39.19 was not read; those standards are behind paywalls or downloads that could not be retrieved, so their content is taken from the standards bodies' own published descriptions and from the iso-thes RDF expression. Specific clause numbers are deliberately not cited.
The ISO 1087:2019 catalogue page blocks automated retrieval, so no ISO 1087 clause text was read; it corroborates the generic/partitive distinction only, and no structural node rests on it alone.
SSSOM (Simple Standard for Sharing Ontological Mappings) and comparable mapping-exchange formats used in life sciences were not researched, so the mapping-set artifact is not aligned to them; adopters in those domains should treat that as an open alignment.
Confidence, similarity scoring and probabilistic mapping justification are not modelled; only a free-text strength basis is carried, which is weaker than what several mapping-exchange formats support.
Faceted classification and citation-order rules beyond node labels and facet indicators are not modelled.
The ISO 25964 XML exchange schema is referenced only through the crosswalk finding; its element-level structure was not enumerated.
No treatment of concept collections as query-time facets, or of collection-driven display ordering beyond the ordered member position.
Local value sets for registration status, change class, obsolescence reason, warrant type, maturity, retention class and disposition are specified as required governed code systems but not enumerated, because their canonical names differ per registration authority.
Signature and non-repudiation mechanics beyond digest recording, and any key management, are omitted; they belong to the platform.
Multilingual governance divergence — where different language communities steward the same concept under different authorities — is not modelled and would need a dedicated stewardship-partition structure.
Machine-actionable expression of the workflow itself, as opposed to the records it emits, is left to the Processes URL referenced from AGENTS.md.

Machine files

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

Provenance

world-models research · reviewable-draft

Built from: models/wm-knw-006-concept-term/spec.yaml, ver-cy/world-models/card-supplements/wm-knw-006-concept-term.json