2.1 Data Models & Standards
How Aevum structures, validates, and standardizes knowledge across disciplines and languages.
Aevum Encyclopedia relies on a strict, interoperable data model to ensure consistency, machine-readability, and cross-lingual alignment. Every entry, entity, and relationship is normalized against established academic and semantic web standards before publication.
Schema.org, Dublin Core, and Wikidata.Core Entity Model
Every piece of content in Aevum is modeled as an Entity with four mandatory dimensions:
- Identity: Unique URI, human-readable label, and Wikidata QID (if applicable).
- Attributes: Type-specific properties (e.g., birth date for persons, coordinates for locations).
- Relationships: Directed edges to other entities (author of, located in, subclass of).
- Provenance: Source citations, revision history, and contributor attribution.
This model enables dynamic knowledge graphs, automated cross-referencing, and real-time fact-checking across our 140+ language editions.
Metadata Standards Compliance
Aevum enforces strict metadata alignment with international standards to guarantee discoverability and interoperability:
| Standard | Application | Requirement |
|---|---|---|
Dublin Core |
Core bibliographic & archival metadata | Mandatory for all entries |
Schema.org |
Web search optimization & type inference | Required for public-facing JSON-LD |
ISO 639-1/3 |
Language tagging | Enforced on all multilingual fields |
Wikidata (QID/PID) |
Canonical entity mapping | Recommended; auto-resolved when available |
ISO 8601 |
Date/time formatting | Strict validation on temporal properties |
JSON-LD Structure Example
Below is a normalized representation of a historical figure entry as ingested by the Aevum pipeline:
{
"@context": "https://schema.org",
"@type": "Person",
"@id": "ae:entity/hypatia_of_alexandria",
"name": {
"en": "Hypatia of Alexandria",
"el": "Ὑπατία Ἀλεξανδρείας",
"la": "Hypatia Alexandrina"
},
"birthDate": "0370-01-01",
"deathDate": "0415-03-08",
"occupation": [
{"@type": "Occupation", "name": "Mathematician"},
{"@type": "Occupation", "name": "Philosopher"},
{"@type": "Occupation", "name": "Astronomer"}
],
"affiliation": {
"@type": "Organization",
"@id": "ae:entity/neoplatonist_school_alexandria",
"name": "Neoplatonist School of Alexandria"
},
"wikidataQID": "Q43802",
"ae:provenance": {
"contributors": ["ae:user/rachel_kim", "ae:user/marcus_wei"],
"lastVerified": "2025-08-12T14:30:00Z",
"sources": [
"https://doi.org/10.1086/344742",
"https://archive.org/details/ancientgreekmath"
]
}
}Validation & Quality Pipeline
Before any data enters the production graph, it passes through a multi-stage pipeline:
- Syntax Check: JSON-LD conformance, required field presence, type validation.
- Ontology Alignment: Automatic mapping to Aevum's controlled vocabulary and Wikidata properties.
- Fact Verification: AI-assisted cross-referencing against trusted academic corpora and primary sources.
- Peer Review Queue: Domain experts validate ambiguous or newly created entities.
- Immutable Publication: Approved entries receive a version hash and are indexed across all language editions.
Contributors receive real-time feedback through the editor interface, highlighting missing provenance, format violations, or conflicting relationships.
Versioning & Immutability
Aevum treats knowledge as an evolving but traceable artifact. Every revision generates a cryptographically signed snapshot:
- Content Hashing: SHA-256 digest of the complete entity state
- Diff Tracking: Machine-readable change logs between versions
- Rollback Capability: Any past version can be restored by approved editors
- Public Ledger: Revision history is publicly accessible and citable