Skip to Content
DocsOntologyData and Ontology

Data and Ontology

Ontology and Mongo data are related but not the same layer.

LayerRole
Ontology (TBox)Vocabulary of kinds: classes, properties, optional SHACL constraints
Data (ABox)Concrete Mongo documents that realize those kinds

Forward: ontology → data

  1. Provide Turtle (upload/paste) in Workbench2 synthetic generation
  2. Backend summarizes classes and properties into prompt context
  3. LLM generates documents preferring ontology terms
  4. Documents insert into the chosen database.collection

Reverse: data → ontology

  1. On Workbench2 Data Management (/data), open a collection and choose Export (download icon)
  2. Set File Type to ttl — indicative ontology (Turtle)
  3. Choose sample size, then Download (or save to the server data folder)
  4. Optionally paste that TTL into Generate Synthetic Data → Ontology for a round trip

Reverse output is inductive: sparse fields, mixed types, and nested arrays may need manual refinement.

Ontology TTL ──guides──► Synthetic generation ──inserts──► Mongo collection
     ▲                                                         │
     └──────── indicative infer (sample) ──────────────────────┘

Runtime asymmetry

For ecosystem-runtime campaigns:

Last updated on
Link TypesAlignments