Classification & Taxonomy
How Aevum Encyclopedia structures, organizes, and interconnects human knowledge through a dynamic, multi-dimensional classification framework.
📐 Overview
Unlike traditional static hierarchies, Aevum's taxonomy is a hybrid graph-hierarchical system. It combines rigid academic disciplines with fluid, AI-augmented semantic relationships, ensuring every article can be reached through multiple contextual paths while maintaining scholarly precision.
Our system is built on three core pillars:
- Disciplinary Roots: Established academic fields as primary anchors
- Faceted Metadata: Tags for themes, eras, regions, methodologies, and audiences
- Dynamic Linking: AI-generated cross-references that evolve with new research
🌳 Hierarchical Structure
The primary classification tree follows a modified Dewey-LCC hybrid, adapted for digital interconnectivity. Expand any node to explore subcategories.
Click any parent node to expand/collapse its branches. Full taxonomy depth reaches 6 levels with ~4.2M leaf nodes.
🏷️ Faceted Classification
To handle interdisciplinary topics, every article is enriched with structured metadata facets. These enable precise filtering and contextual discovery.
Temporal Facet
Historical periods, eras, or date ranges (e.g., "20th Century", "Late Bronze Age")
Geographic Facet
Regions, countries, biomes, or cosmological scopes tied to the subject
Methodology Facet
Research approaches, theoretical frameworks, or artistic movements
Audience Facet
Reading level & specialization tier (Undergrad → Graduate → Expert)
Facets are stored as structured JSON-LD and exposed via our search API for programmatic filtering.
🔍 Methodology & Curation
Classification is not automatic. It follows a rigorous human-in-the-loop pipeline:
- Initial AI Suggestion: NLP models propose primary category + 3–5 facets based on text analysis
- Editorial Review: Subject-matter editors validate placement and adjust for nuance
- Peer Cross-Check: A second reviewer from a related discipline ensures no siloing
- Graph Integration: Approved taxonomy entries are linked to the knowledge graph via RDF triples
- Quarterly Audit: ML drift and academic paradigm shifts trigger reclassification reviews
This ensures that taxonomy remains academically rigorous while adapting to emerging fields.
🔗 Cross-References & Knowledge Mapping
Articles are never isolated. Our semantic engine maintains three types of relational mapping:
- Prerequisite Links: Foundational concepts required to understand the current topic
- Contrastive Links: Related but opposing theories, movements, or classifications
- Application Links: Practical implementations, case studies, or downstream effects
These relationships power the interactive Knowledge Graph visualization and influence search ranking weights.
⚙️ Developer Resources
Access the full taxonomy tree, facet dictionaries, and relationship endpoints via our REST & GraphQL APIs.
Schema Example (JSON-LD)
API Endpoints
GET /api/v3/taxonomy/tree— Retrieve full hierarchical structureGET /api/v3/taxonomy/facets— Query available metadata dimensionsPOST /api/v3/taxonomy/classify— Submit content for AI classification review
Full documentation, SDKs, and rate limits are available in the Developer Portal.