Semantic Web Domain Ontology

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

FLOW "SemanticWebDomainOntology" v1.0

# PURPOSE
Represent linked, machine-interpretable web data with explicit semantics, verifiable provenance, and bounded inference.

# TYPED ENTITIES
ENTITY WebResource { iri: IRI, label: String?, type: IRI* }
ENTITY Dataset { iri: IRI, publisher: IRI?, licence: IRI?, version: String? }
ENTITY RDFTriple { subject: IRI|BlankNode, predicate: IRI, object: IRI|BlankNode|Literal }
ENTITY NamedGraph { iri: IRI, triples: RDFTriple*, provenance: Provenance? }
ENTITY Ontology { iri: IRI, classes: IRI*, properties: IRI*, versionIRI: IRI? }
ENTITY VocabularyTerm { iri: IRI, definition: String?, language: BCP47? }
ENTITY Class { iri: IRI, superclass: IRI* }
ENTITY Property { iri: IRI, domain: IRI*, range: IRI*, subpropertyOf: IRI* }
ENTITY Literal { lexicalForm: String, datatype: IRI?, language: BCP47? }
ENTITY Linkset { sourceDataset: IRI, targetDataset: IRI, mappings: Mapping* }
ENTITY Mapping { source: IRI, target: IRI, relation: MappingRelation, evidence: String? }
ENTITY Shape { iri: IRI, target: IRI, constraints: Constraint* }
ENTITY Query { language: "SPARQL", expression: String }
ENTITY Provenance { source: IRI?, agent: IRI?, generatedAt: DateTime?, method: String? }
ENTITY Assertion { triple: RDFTriple, graph: IRI?, confidence: Decimal?, provenance: Provenance? }

# RELATIONSHIPS
RELATION hasType(WebResource, Class)
RELATION subclassOf(Class, Class)
RELATION subpropertyOf(Property, Property)
RELATION equivalentClass(Class, Class)
RELATION equivalentProperty(Property, Property)
RELATION sameAs(WebResource, WebResource)
RELATION seeAlso(WebResource, WebResource)
RELATION definedBy(VocabularyTerm, Ontology)
RELATION describes(Dataset, WebResource)
RELATION containsTriple(NamedGraph, RDFTriple)
RELATION conformsTo(Dataset, Shape)
RELATION derivedFrom(Assertion, Dataset)
RELATION linkedTo(WebResource, WebResource)

# SEMANTIC RULES
RULE R1: IF subclassOf(A,B) AND hasType(x,A) THEN infer hasType(x,B).
RULE R2: IF subpropertyOf(p,q) AND triple(x,p,y) THEN infer triple(x,q,y).
RULE R3: IF equivalentClass(A,B) THEN class membership may propagate in both directions under the chosen OWL entailment regime.
RULE R4: IF equivalentProperty(p,q) THEN property assertions may propagate in both directions under the chosen OWL entailment regime.
RULE R5: IF sameAs(x,y) is accepted under OWL semantics THEN corresponding identity substitution is permitted; do not infer sameAs from label similarity alone.
RULE R6: IF a shape validation fails THEN report a data-quality violation; do not equate SHACL nonconformance with logical inconsistency.
RULE R7: IF a SPARQL query runs THEN record dataset snapshot, entailment regime, and query version when reproducibility is required.
RULE R8: IF a source assertion conflicts with another THEN retain both with provenance until a specified conflict-resolution policy applies.

# CONSTRAINTS
CONSTRAINT C1: IRIs identify resources; literals represent lexical values with datatype or language tags.
CONSTRAINT C2: RDF blank nodes are scoped identifiers, not globally persistent IRIs.
CONSTRAINT C3: RDF/OWL open-world reasoning does not treat absence of a triple as falsity.
CONSTRAINT C4: RDF graphs are sets of triples; provenance may be attached through named graphs or assertion records.
CONSTRAINT C5: OWL reasoning, SHACL validation, and SPARQL querying are distinct operations.
CONSTRAINT C6: Identity links require stronger evidence than topical similarity or lexical equivalence.
CONSTRAINT C7: External vocabulary reuse should preserve source definitions, licensing, and version references.

# CONFIDENCE
CONFIDENCE.range: [0.0, 1.0]
CONFIDENCE.applies_to: [extracted_assertion, entity_link, mapping, inferred_assertion]
CONFIDENCE.rule: Explicit source assertions and logically entailed statements are distinguished from uncertain extraction and heuristic mappings.
CONFIDENCE.propagation: Do not automatically multiply confidence scores or interpret logical entailment as empirical certainty.

# SEMANTIC_EQUIVALENCE
SEMANTIC_EQUIVALENCE.levels: [exact_identity, ontology_equivalence, close_match, related_match, lexical_similarity]
SEMANTIC_EQUIVALENCE.policy: exact_identity requires justified identity; close_match and related_match never imply owl:sameAs.
SEMANTIC_EQUIVALENCE.context: [language, domain, ontology_version, dataset, temporal_scope]

# PROVENANCE
PROVENANCE.capture: [source_iri, publisher, retrieval_time, named_graph, extraction_method, vocabulary_version, licence]
PROVENANCE.rule: Derived assertions retain links to supporting assertions and the inference rule applied.

# FLOW_HASH
FLOW_HASH.algorithm: SHA-256
FLOW_HASH.input: Canonical UTF-8 serialization of flow definition, excluding the computed hash field.
FLOW_HASH.note: Hash must be calculated at deployment; no hash value is fabricated here.

# UPDATE_MODEL
UPDATE_MODEL.strategy: versioned_incremental
UPDATE_MODEL.steps: [ingest, normalise, validate, reconcile, infer, index, publish]
UPDATE_MODEL.on_change: Revalidate affected shapes, refresh dependent inferences, preserve prior versions and provenance.
UPDATE_MODEL.conflict_policy: flag_and_review
UPDATE_MODEL.rollback: restore_previous_version

# AI_CONSUMPTION
AI_CONSUMPTION.format: [RDF_Turtle, JSON_LD, RDF_NQuads, SPARQL_results]
AI_CONSUMPTION.entrypoints: [ontology_iri, dataset_iri, named_graph_iri]
AI_CONSUMPTION.require: [typed_entities, explicit_relations, provenance, inference_regime, confidence_annotations]
AI_CONSUMPTION.guardrails: [do_not_invent_triples, distinguish_asserted_from_inferred, respect_licences, retain_context, avoid_unjustified_identity]
AI_CONSUMPTION.output: machine-readable facts with traceable sources and declared uncertainty.

END FLOW