Semantic Web & Knowledge Representation

The Inquiry: The Semantic Web provides mature, standardized infrastructure for representing, reasoning over, and querying structured knowledge — OWL 2 (2012), RDF 1.1 (2014), PROV-O (2013), SKOS (2009), SPARQL (2013), OWL-Time (2017). Foundational ontologies (BFO, DOLCE, SUMO, GFO) have demonstrated that domain-independent upper ontologies can integrate hundreds of domain ontologies at scale (BFO: hundreds of biomedical ontologies, ISO/IEC 21838 standardization). Yet no existing ontology, knowledge graph architecture, or semantic framework addresses organizational decision governance — capturing why decisions were made, under what authority, with what evidence, subject to which constraints, and how decisions relate to the governance chain. Is this an oversight (solvable by applying existing standards) or a genuine application gap requiring novel ontological contribution?

Falsifiable formulation: If any existing ontology, knowledge graph, or semantic framework provides the formal infrastructure to capture, reason over, and validate organizational decision governance — including authority traceability, evidence provenance, constraint compliance, and temporal validity — then the application gap claimed here does not exist.

The Semantic Web has solved the technical infrastructure problem; organizational decision governance is the unsolved application.

Three decades of W3C standardization, description logic research, and foundational ontology work have produced remarkably mature infrastructure. OWL 2 provides decidable reasoning with computational guarantees. RDF provides a universal data model. PROV-O provides provenance tracking. SKOS provides vocabulary management. SPARQL provides query infrastructure. BFO demonstrates that upper ontologies can integrate hundreds of domain ontologies at scale (through OBO Foundry). Description logic research has fully characterized the tractability-expressiveness trade-off. Ontology engineering methodologies (METHONTOLOGY, NeOn, competency questions, design patterns) provide validated construction approaches.

The application gap is specific: no existing ontology captures the semantics of organizational decision governance. Enterprise knowledge graphs model entities for query and discovery. PROV-O tracks provenance. Business process ontologies capture workflows. But none captures why a decision was made, under what authority, with what evidence, subject to which constraints, and how it relates to the governance chain. This gap persists not because the technical infrastructure is insufficient but because the domain has been modeled procedurally (BPM, workflow systems) rather than ontologically (formal KR with decidable reasoning).


Knowledge representation is not neutral — it constitutes a theory of governance.

Davis, Shrobe & Szolovits's (1993) five-role framework reveals that a governance ontology is not merely a data format but a theory of what governance means. The choice of primitives determines what the system can "see" (ontological commitments). The sanctioned inferences define what counts as valid governance reasoning. The computational profile determines what governance questions can be answered in bounded time. The human expression requirement ensures that governance conclusions remain interpretable. Gruber's (1993) ontological commitment concept adds: adopting a governance ontology is a social contract — organizations agree to represent governance using shared vocabulary, enabling interoperability not through data format conversion but through shared conceptualization.


The passive-to-active transition: from knowledge graph to governance substrate.

Enterprise knowledge graphs and Graph RAG represent the current state of the art in organizational knowledge infrastructure — but they are passive. They represent what is, not what should be. They answer "what happened?" not "was it authorized?" The transition from passive knowledge graph to active governance substrate — where the ontology doesn't just describe governance but constrains, validates, and enables reasoning over governance actions — represents the novel application of mature semantic web infrastructure to an underserved domain.