Positioning Neuro-Symbolic AI in Domain Engineering

Documentation status: guide — see Maturity and evidence.

Neuro-symbolic domain engineering combines neural systems for interpretation and proposal generation with explicit symbolic structures for identity, constraints, reasoning, and validation.

Correctly positioned use

A neural model may extract a candidate relation from documentation. The symbolic layer then checks whether the participant concepts exist, whether their ontology levels are compatible, whether role names are valid, and whether the relation violates declared constraints.

The neural component proposes and generalizes. The symbolic component structures and verifies.

Principle

Source material
  -> neural extraction / candidate generation
  -> symbolic representation
  -> constraint evaluation
  -> review / correction
  -> accepted conceptual model