Relationships

Concepts in Glossarist are connected by typed semantic relationships — directional links that express how one concept relates to another. Every relationship has a type, a source concept, and a target concept. See the RelatedConcept entity definition for the authoritative field list.

This page documents all relationship types organized by category, their ISO/SKOS alignment, and how to author them in YAML. The relationship definitions are derived from the Glossarist concept model ontology.

Overview

Relationships serve several purposes:

• Navigation — Users traverse concept systems by following relationships (e.g., "narrower" to drill into specifics)
• Reasoning — Hierarchical relationships enable inheritance of properties across the concept tree
• Interoperability — Standard relationship types map to SKOS and ISO 25964 properties for cross-vocabulary exchange
• Provenance — Lifecycle relationships track how concepts evolve over time (superseded, deprecated)
• Cross-dataset linking — Relationships connect concepts across datasets and editions (e.g., VIM editions, G18 to VIM/VIML)

Each relationship is directional: A --[type]--> B means "A has relationship type to B." Many types have an inverse: if A broader B, then B narrower A.

Relationship Types

Loading relationship types…

YAML Authoring

Relationships are expressed in concept YAML files under the top-level related key. Each entry has a type and a content reference to the target concept.

Hierarchical relationships

termid: "3.1.1.1"
status: valid
eng:
  terms:
    - type: expression
      designation: "entity"
      normative_status: preferred
  definition:
    - content: "A concrete or abstract thing"
  sources:
    - type: authoritative
      origin: "ISO 19107:2003"

# Relationships to other concepts
related:
  - type: broader
    content: "3.1.1"          # This concept is narrower than 3.1.1
  - type: narrower
    content: "3.1.1.1.1"      # This concept is broader than 3.1.1.1.1
  - type: broader_generic
    content: "3.1.1"          # BTG: this is a kind of 3.1.1
  - type: narrower_partitive
    content: "3.1.1.1.2"      # NTP: 3.1.1.1.2 is a part of this

Equivalence and mapping

related:
  - type: exact_match
    content: "iso-19107:entity"   # Exact match in another vocabulary
  - type: broad_match
    content: "iso-19109:object"   # Broader concept in another vocabulary
  - type: equivalent
    content: "3.1.1.1"            # Equivalent concept in same vocabulary

Associative and lifecycle

related:
  - type: compare
    content: "3.1.1.2"            # Compare with another concept
  - type: see
    content: "3.1.2"              # See also
  - type: deprecates
    content: "3.1.1.0"            # This concept deprecates 3.1.1.0
  - type: supersedes
    content: "3.1.1.0"            # This concept supersedes 3.1.1.0

Cross-dataset relationships

When linking to a concept in a different dataset, use the ref object with a source URN and concept id:

related:
  - type: supersedes
    ref:
      source: urn:oiml:pub:v:1:2013
      id: '2.1'
  - type: see
    ref:
      source: urn:oiml:pub:v:2-200:2012
      id: '2.13'

The source URN identifies the target dataset, and id is the concept identifier within that dataset. The concept-browser resolves these at build time into navigable cross-dataset edges.

Multiple relationships

A concept can have many relationships of different types:

related:
  - type: broader_generic
    content: "3.1"
  - type: narrower_generic
    content: "3.1.1.1"
  - type: related_concept
    content: "3.2.5"
  - type: compare
    content: "3.1.2"
  - type: exact_match
    content: "iso-tc211:entity"

Designation-level relationships

Designation-level relationships (abbreviated_form_for, short_form_for) link designations of the same concept to each other. They appear on designations rather than concepts, and use target (designation text) instead of ref (concept identifier).

terms:
  - type: expression
    designation: Light Emitting Diode
    normative_status: preferred
  - type: abbreviation
    designation: LED
    normative_status: admitted
    acronym: true
    related:
      - type: abbreviated_form_for
        target: Light Emitting Diode

See Designations for the full DesignationRelationship model. The ref.text field on concept-level relationships serves a different purpose — it provides designation text when the concept is identified by text rather than by registry reference.

Inverse Derivation

Glossarist derives inverse relationships automatically from the relationship graph. When concept A declares supersedes concept B, the system derives superseded_by on concept B at render time — without requiring the inverse to be explicitly authored.

This means you only author the forward direction. Inverse relationships appear in the concept-browser automatically:

Authored type	Derived inverse
supersedes	superseded_by
deprecates	deprecated_by
replaces	replaced_by
invalidates	invalidated_by
retires	retired_by
broader	narrower
has_part	is_part_of
instance_of	has_instance
has_version	version_of

Symmetric types (equivalent, compare, contrast, close_match, related_match, related_concept) are their own inverse — no derivation needed.

Cross-Dataset Navigation

Glossarist's concept-browser supports cross-dataset relationships — links between concepts in different datasets loaded from the same deployment. This enables multi-edition navigation for terminology systems like OIML VIM/VIML.

How it works

1. Each dataset defines concepts with unique URNs (e.g., urn:oiml:pub:v:1:2022)
2. Concepts declare related entries using ref.source to target a different dataset
3. At build time, the concept-browser resolves all cross-dataset references into an edge graph
4. The UI renders these as navigable links with category-colored badges

Example: VIM multi-edition navigation

The OIML VIM (International Vocabulary of Metrology) has four editions (1993, 2007, 2010, 2012). Each edition is a separate dataset. Concepts in newer editions declare supersedes relationships to their predecessors:

# In datasets/vim-2012/concepts/2.13.yaml
related:
  - type: supersedes
    ref:
      source: urn:oiml:pub:v:2-200:2007
      id: '2.13'

When rendered, the VIM 2012 concept 2.13 shows a "supersedes" link to VIM 2007 concept 2.13, and the VIM 2007 concept automatically shows "superseded by" pointing back.

Example: G18 cross-references to VIM/VIML

OIML G 18 is an alphabetical glossary of 2132 terms sourced from 88 OIML publications. Many terms originate from VIM or VIML. Cross-dataset see relationships link G18 terms to their authoritative definitions:

# In datasets/g18/concepts/00079.yaml (measuring instrument)
related:
  - type: see
    ref:
      source: urn:oiml:pub:v:2-200:2007
      id: '3.1'

This enables users navigating G18 to jump directly to the authoritative VIM definition with full notes and examples.

SKOS Alignment

Glossarist relationship types map to standard SKOS and ISO 25964 ontology properties:

Glossarist type	SKOS / ISO 25964 property	Notes
broader	skos:broader	Generic broader
narrower	skos:narrower	Generic narrower
broader_generic	iso-thes:broaderGeneric	ISO 25964 BTG
narrower_generic	iso-thes:narrowerGeneric	ISO 25964 NTG
broader_partitive	iso-thes:broaderPartitive	ISO 25964 BTP
narrower_partitive	iso-thes:narrowerPartitive	ISO 25964 NTP
broader_instantial	iso-thes:broaderInstantial	ISO 25964 BTI
narrower_instantial	iso-thes:narrowerInstantial	ISO 25964 NTI
exact_match	skos:exactMatch	Cross-vocabulary
close_match	skos:closeMatch	Cross-vocabulary
broad_match	skos:broadMatch	Cross-vocabulary
narrow_match	skos:narrowMatch	Cross-vocabulary
related_match	skos:relatedMatch	Cross-vocabulary
related_concept	skos:semanticRelation	General associative
equivalent	skos:exactMatch	Within-vocabulary

Types without a direct SKOS equivalent (compare, contrast, see, deprecates, supersedes, homograph, false_friend, ISO 19135 register management types, etc.) are expressed as gloss: properties in the Glossarist OWL ontology and serialized as-is in TBX output using the corresponding TBX data category.

Design Principles

1. MECE coverage — The relationship types form a mutually exclusive, collectively exhaustive set covering all ISO 10241-1, ISO 19135, ISO 25964, SKOS, ISO 12620, and TBX relationship categories
2. No overlap with SKOS — Where SKOS defines a property, Glossarist reuses the exact same semantics (e.g., broader = skos:broader)
3. Directional — Every relationship has a source and target; inverse pairs are separate types (broader/narrower, supersedes/superseded_by)
4. Typed hierarchy — The three ISO 25964 hierarchy types (generic, partitive, instantial) are distinct because they carry different semantic meaning and affect how concept trees are rendered
5. Cross-dataset by design — Relationships use URN-based references that span datasets, enabling multi-edition and cross-vocabulary navigation

See the YAML Schema Reference for the JSON Schema definition of the RelatedConcept entity, the Entity Field Reference for field-level details, and the Standards compliance reference for ISO standard mappings.