Semantic Ontology Framework

The formal knowledge representation layer that powers Aevum Encyclopedia's reasoning engine, enabling machine-readable relationships, cross-domain inference, and precise semantic search across 2.4M+ articles.

Overview

The Aevum Semantic Ontology is a rigorously structured, OWL 2 DL-compliant knowledge schema designed to model entities, concepts, relationships, and constraints across all domains of human knowledge. Unlike flat taxonomies, our ontology supports multi-parent inheritance, property restrictions, and logical axioms that enable automated reasoning and disambiguation.

ℹ️

This ontology serves as the single source of truth for entity resolution, citation mapping, and AI-assisted content structuring across the Aevum platform.

Ontology Architecture

The framework operates on a three-tier architecture balancing expressivity, performance, and extensibility:

Aevum Core Ontology (ACO)

Domain Modules

Cross-Domain Bridges

Temporal/Geospatial Extensions

Entity Classes

Object/DataType Properties

Axioms & Constraints

Each tier is independently versioned but jointly validated to maintain DL-compliance. Domain modules (e.g., aevo:biology, aevo:history) inherit from the core schema while introducing specialized property domains and ranges.

Core Components

Classes Entity Taxonomy

Structured hierarchy including aevo:Concept, aevo:PhysicalEntity, aevo:Event, aevo:AbstractWork, and aevo:Agency with strict disjointness axioms.

Properties Relationship Schema

Object properties (e.g., aevo:partOf, aevo:derivesFrom) and datatype properties (e.g., aevo:hasCitationDOI, aevo:temporalExtent) with defined domains, ranges, and cardinality constraints.

Axioms Logical Constraints

SWRL rules and OWL restrictions enforcing consistency: e.g., aevo:Event must have exactly one aevo:temporalExtent, and aevo:Person is disjoint with aevo:Organization.

Property Classification Table

Property Type	Prefix	Example	Cardinality
Object Property	aevo:rel	aevo:rel:causes, aevo:rel:influences	0..*, 1..1, 0..1
Datatype Property	aevo:attr	aevo:attr:hasISOCode, aevo:attr:publicationYear	0..1, 1..1
Annotation Property	aevo:meta	aevo:meta:editorialNote, aevo:meta:confidenceScore	Unbounded

Technical Implementation

The ontology is published as a modular OWL 2 DL knowledge base, serialized in Turtle and RDF/XML, with real-time SPARQL 1.1 query access. All triples are stored in a distributed Triplestore optimized for pattern matching and inference.

Turtle Serialization OWL 2 DL

@prefix aevo: <http://ontology.aevum.org/core/#> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .

aevo:Concept a owl:Class ;
    rdfs:label "Concept" @en ;
    rdfs:comment "Abstract intellectual entity representing a defined notion." @en .

aevo:rel:derivesFrom a owl:ObjectProperty ;
    rdfs:domain aevo:AbstractWork ;
    rdfs:range aevo:AbstractWork ;
    owl:propertyChainAxiom (aevo:rel:cites aevo:rel:founds) .

⚠️

DL compliance is enforced via automated reasoning checks. Any submitted ontology extension must pass HermiT/TBox validation before merging into the production graph.

Querying the Graph

Researchers and developers can query the ontology using standard SPARQL 1.1. The public endpoint supports federated queries, graph patterns, and constructed results for downstream integration.

SPARQL Example Endpoint

SELECT ?entity ?label ?related ?confidence
WHERE {
  ?entity a aevo:ScientificTheory ;
          rdfs:label ?label .
  OPTIONAL {
    ?related aevo:rel:contradicts ?entity .
    ?entity aevo:meta:confidenceScore ?confidence .
  }
}
ORDER BY DESC(?confidence)

The query above retrieves scientific theories alongside known contradictions and their editorial confidence scores, demonstrating how ontological constraints enable precise, context-aware retrieval.

Versioning & Governance

Ontology evolution follows a strict lifecycle:

Draft: Proposal reviewed by domain ontology stewards
Validation: Automated consistency checks + manual peer review
Release: Semantic versioning (vMAJOR.MINOR.PATCH) with deprecation notices for breaking changes
Alignment: Quarterly mapping audits against Schema.org, Wikidata, and DBpedia

Deprecations are handled via owl:deprecated annotations with migration paths. Backward compatibility is preserved through property chaining and synonym mapping.

Standards Alignment

Aevum's ontology maintains formal alignments with major knowledge infrastructure projects to enable interoperability:

Standard	Alignment Method	Scope
Schema.org	owl:equivalentClass / rdfs:subClassOf	General entities, creative works, events
Wikidata (WD)	owl:sameAs / SKOS mapping	Entity resolution, cross-lingual labels
DCAT / PROV-O	Property chaining	Data cataloging, provenance tracking