Ontology & Vocabularies
The formal semantic layer that gives the knowledge graph its meaning — OWL classes and properties, SKOS controlled vocabularies with alternative labels for data harmonization, and provenance metadata for full data lineage.
Data Lineage
Every node carries provenance metadata for traceability back to the original data source.
_source
Identifies the data origin — e.g. ishares_api, kaggle_esg, yfinance.
_ingested_at
ISO 8601 timestamp of when this entity was written to the graph.
_pipeline_run_id
Unique identifier for the ETL pipeline execution that created or updated this node.
Example: Portfolio SWDA
RDF Export
Export the Neo4j graph as RDF triples using the AMKG ontology namespace. Each node becomes a typed resource with datatype properties mapped to their OWL definitions.
URI pattern: amkg:{Label}_{primary_key} — e.g. amkg:Portfolio_SWDA, amkg:Asset_IE00B4L5Y983
SPARQL Query
Query the knowledge graph using SPARQL. The graph is exported to RDF in-memory and queried with rdflib — no external triplestore required.