OWL Ontology¶

Own the ontology. Own the perspective. The OWL ontology is what transforms raw data into company-specific knowledge.

What the Ontology Does¶

The OWL ontology (knowledge_graph/ontology.ttl, ~26KB) defines the company-specific perspective that transforms raw data into organizational knowledge. It is not a static schema — the OWLBridge runs active reasoning cycles that discover new facts through logical inference.

Standards Alignment¶

Standard	Purpose	How We Use It
BFO (Basic Formal Ontology)	Upper ontology for foundational categories	Root classes for Entity, Process, Quality
PROV-O (W3C Provenance)	Attribution and derivation tracking	`wasGeneratedBy`, `wasDerivedFrom`, `wasAttributedTo` edges
SKOS (Knowledge Organization)	Taxonomic hierarchies	`broader`, `narrower`, `relatedConcept` for concept navigation
FIBO (Financial Industry)	Financial domain concepts	`FinancialInstrument`, `Account`, `Transaction` classes

Reasoning Cycle: Promote → Reason → Downfeed¶

The OWLBridge operates a three-phase cycle:

1. Promote (LPG → OWL)¶

High-importance nodes from the property graph (NetworkX/LadybugDB) are promoted into the OWL ontology as named individuals:

NetworkX Node: {"id": "customer:001", "risk_level": "high", "importance_score": 0.9}
    ↓ promote
OWL Individual: :Customer001 rdf:type :Customer ; :hasRiskLevel "high"

2. Reason (OWL Inferencing)¶

The OWL reasoner applies: - Transitive closure — If A dependsOn B and B dependsOn C, then A dependsOn C - Symmetric properties — If A relatedTo B, then B relatedTo A - RDFS+ entailment — Subclass hierarchies propagate properties downward - SKOS hierarchy — broader/narrower relationships enable category navigation

3. Downfeed (OWL → LPG)¶

Newly inferred facts are written back to the property graph as edges:

OWL Inference: :Customer001 :dependsOn :Service002 (via transitive closure)
    ↓ downfeed
NetworkX Edge: ("customer:001", "service:002", "depends_on")
    + {"inferred": true, "inference_type": "transitive_closure"}

Ontological Lensing¶

The same underlying data can be viewed through different ontological perspectives — the "different lenses" concept:

Risk Lens — View the graph through risk propagation edges
Compliance Lens — View through regulatory and audit edges
Operational Lens — View through dependency and process edges
Financial Lens — View through FIBO-aligned financial edges

Key OWL Classes (from ontology.ttl)¶

Class	Description	BFO Alignment
`:Agent`	Any actor (human or AI)	`bfo:IndependentContinuant` (`bfo:0000004`)
`:Episode`	A discrete interaction unit / bounded work session	`bfo:Process` (`bfo:0000015`)
`:Concept`	An abstract knowledge unit	`bfo:GenericallyDependentContinuant`
`:Decision`	A recorded decision with rationale	`bfo:Process`
`:Policy`	An organizational rule or constraint	`bfo:GenericallyDependentContinuant`
`:Skill`	A reusable agent capability	`bfo:IndependentContinuant` (`bfo:0000004`)
`:Evidence`	A claim or finding grounded in source material	`bfo:GenericallyDependentContinuant` (`bfo:0000031`)

Company-Specific Extension¶

To add your own domain concepts, extend ontology.ttl:

:CustomerSegment rdfs:subClassOf bfo:GenericallyDependentContinuant ;
    rdfs:label "Customer Segment" ;
    rdfs:comment "A company-specific classification of customers." .

:belongsToSegment rdf:type owl:ObjectProperty ;
    rdfs:domain :Customer ;
    rdfs:range :CustomerSegment ;
    rdfs:label "belongs to segment" .

The OWLBridge will automatically include new classes and properties in its reasoning cycle without any code changes.