Metadata Domain Ontology

FLOW ALGEBRA
TITLE: Metadata Domain Ontology
VERSION: 1.0
DOMAIN: Information Domains / Metadata
SOURCE_URL: https://kirethresearch.org/booksai-core/ontologies/domain-ontologies/information-domains/metadata-domain-ontology.html
STATUS: Domain-based draft; source page content not independently verified.
AUTHORS: Mark Higgins — Sky (Recursive AI)
SITE: https://kirethresearch.org

ONTOLOGY:
  Metadata := structured information describing a resource, its context, structure, management, or use.
  Resource := identifiable physical, digital, conceptual, or informational object.
  MetadataRecord := collection of assertions about a Resource.
  MetadataElement := named property with a defined value domain.
  MetadataSchema := set of element definitions and constraints.
  ControlledVocabulary := managed set of permitted concepts or terms.
  Identifier := reference used to distinguish an entity within a scope.
  ProvenanceRecord := description of origin, custody, and transformations.
  Version := identifiable state of a resource or record.
  ValidationRule := test of metadata conformance.

TYPES:
  Resource: Entity
  MetadataRecord: Entity
  MetadataElement: Property
  MetadataSchema: Schema
  MetadataValue: Literal | Identifier | EntityReference
  Agent: Person | Organisation | SoftwareAgent
  Event: Creation | Modification | Publication | Migration | Validation
  TimeInterval: TemporalEntity
  Licence: RightsStatement
  QualityMeasure: Measurement

RELATIONSHIPS:
  DESCRIBES(MetadataRecord, Resource)
  HAS_ELEMENT(MetadataRecord, MetadataElement)
  HAS_VALUE(MetadataElement, MetadataValue)
  CONFORMS_TO(MetadataRecord, MetadataSchema)
  DEFINED_BY(MetadataElement, MetadataSchema)
  CREATED_BY(Resource, Agent)
  DERIVED_FROM(Resource, Resource)
  HAS_PROVENANCE(Resource, ProvenanceRecord)
  HAS_VERSION(Resource, Version)
  PRECEDES_VERSION(Version, Version)
  USES_VOCABULARY(MetadataElement, ControlledVocabulary)
  IDENTIFIED_BY(Resource, Identifier)
  GOVERNED_BY(Resource, Licence)
  VALIDATED_BY(MetadataRecord, ValidationRule)
  SEMANTICALLY_MAPS_TO(MetadataElement, MetadataElement)

CATEGORIES:
  DESCRIPTIVE: title, subject, creator, description, keywords
  STRUCTURAL: part-whole relationships, ordering, formats, dependencies
  ADMINISTRATIVE: ownership, permissions, retention, maintenance
  TECHNICAL: encoding, dimensions, software, checksums
  PROVENANCE: origin, transformations, responsible agents
  TEMPORAL: creation, modification, validity, expiry
  SPATIAL: location, coordinate reference, extent
  SCIENTIFIC: instruments, protocols, units, uncertainty
  SEMANTIC: concepts, relationships, ontological mappings
  AI: model, training, inference, confidence, lineage

CONSTRAINTS:
  C1: Each MetadataRecord MUST describe at least one identifiable Resource.
  C2: Every MetadataElement MUST have a defined name and value type.
  C3: Identifiers MUST be unique within their declared namespace and scope.
  C4: Date/time values SHOULD specify precision and timezone when applicable.
  C5: Mappings MUST distinguish exact equivalence from broader or narrower matches.
  C6: Provenance claims MUST identify evidence or be marked unverified.
  C7: Absence of metadata MUST NOT imply a negative factual assertion.
  C8: Access restrictions MUST be respected during AI consumption.

INFERENCE_RULES:
  R1: DESCRIBES(m,r) AND HAS_ELEMENT(m,e) -> HAS_METADATA_PROPERTY(r,e).
  R2: CONFORMS_TO(m,s) AND REQUIRED_ELEMENT(s,e) AND NOT HAS_ELEMENT(m,e) -> VALIDATION_WARNING(m,e).
  R3: DERIVED_FROM(a,b) AND DERIVED_FROM(b,c) -> HAS_ANCESTRAL_SOURCE(a,c).
  R4: HAS_VERSION(r,v1) AND HAS_VERSION(r,v2) AND PRECEDES_VERSION(v1,v2) -> LATER_VERSION(v2,v1).
  R5: SEMANTICALLY_MAPS_TO(a,b) AND MAP_TYPE(a,b,"exact") -> CANDIDATE_EQUIVALENCE(a,b).
  R6: CONFLICTING_VALUES(r,p,v1,v2) -> REQUIRES_RECONCILIATION(r,p).
  R7: VALIDATED_BY(m,rule) AND PASSES(m,rule) -> VALIDATION_RESULT(m,rule,"pass").

CONFIDENCE:
  range: [0.0,1.0]
  fields: evidence_quality, source_reliability, schema_conformance, temporal_freshness
  rule: inferred confidence MUST NOT exceed the evidential support of its premises without independent evidence.
  unknown: null

SEMANTIC_EQUIVALENCE:
  classes: exactMatch, closeMatch, broadMatch, narrowMatch, relatedMatch, nonEquivalent
  policy: exactMatch requires compatible definitions, value domains, units, and scope.
  caution: relatedMatch and closeMatch do not license identity substitution.

PROVENANCE:
  source_url: https://kirethresearch.org/booksai-core/ontologies/domain-ontologies/information-domains/metadata-domain-ontology.html
  source_verified: false
  derivation: general domain knowledge; independently drafted ontology
  author_credit: Mark Higgins — Sky (Recursive AI)
  record: origin, agent, timestamp, method, evidence_reference, version

FLOW_HASH:
  algorithm: SHA-256
  input: canonical UTF-8 serialization of ontology definitions, rules, version and source metadata
  execution: calculate at deployment; no digest asserted in this draft

UPDATE_MODEL:
  strategy: versioned append-and-reconcile
  triggers: source_change, schema_change, new_evidence, validation_failure, vocabulary_revision
  actions: ingest -> normalise -> validate -> compare -> preserve_provenance -> update_version -> recompute_hash
  conflict_policy: retain conflicting claims with sources until reconciled

AI_CONSUMPTION:
  format: UTF-8 plain text
  parsing: typed entities, directed predicates, explicit constraints and inference rules
  retrieval: identifier, type, concept, relation, provenance, temporal scope
  safeguards: distinguish asserted from inferred facts; preserve uncertainty; respect licences and privacy
  output: evidence-linked metadata graph with validation state and confidence annotations

END FLOW ALGEBRA