Ontology Graph · Neuro-Symbolic Curriculum Engine

Click a node to inspect · double-click to build its prerequisite path · drag to rearrange

0 of 0 relations cross a subject boundary

No concept selected Pick any node in the graph to read its class, grade band, and every asserted triple it takes part in.

Why this shape of data

Every node carries a class (Concept, Law, Equation, Quantity, Process, Structure, Method). Every edge carries a verb with a declared domain and range. Prerequisites are just one verb among several, which is why the graph can answer “what must come first?” and “is this claim even well-formed?” with the same structure.

Target concept

Teach this Learner already knows

Topological sort over the transitive closure of requires. Cycles are rejected at authoring time.

No path builtChoose a target concept, the engine returns the ordered, minimal sequence and counts every subject boundary it has to cross.

How grounding works

NeuralLLM drafts

→

SymbolicGraph checks

→

GroundedCited answer

The model proposes triples. The ontology decides whether they survive: does the object exist, do the classes satisfy the verb's domain and range, and is the relation asserted or entailed by a path? Nothing unprovable reaches the learner.

Model output, unverified

Assert your own triple

Subject Verb Object

Same schema, other domains

Primitive	Education	Legal	Delivery
Class	Concept, Law, Equation	Statute, Clause, Precedent	Task, Deliverable, Risk
requires	Prerequisite topic	Condition precedent	Finish-to-start dependency
governs	Law → phenomenon	Statute → conduct	Policy → process
isFormalizedBy	Concept → equation	Principle → section	Objective → KPI
Constraint	No cycles, grade band	Jurisdiction, in-force date	No circular deps, capacity
Catches	Invented prerequisite	Fabricated citation	Impossible schedule