Semantic Interoperability & Ontology

Semantic interoperability ensures that data exchanged between systems retains its meaning regardless of the source format or originating system. SDL achieves this through standards-based ontology alignment, virtual semantic mappings, and a knowledge graph projection layer that lets users query heterogeneous data as though it resided in a single, unified model.

BFO/CCO Compliance

SDL’s semantic layer is built on Basic Formal Ontology (BFO) and Common Core Ontologies (CCO), in compliance with the DNI memorandum mandating their adoption across the defense and intelligence communities.

  • Shared top-level framework — BFO provides a domain-neutral upper ontology that defines the most general categories of entities and processes. CCO extends BFO with mid-level concepts for information, agents, events, and artifacts.

  • Cross-organizational reasoning — because BFO and CCO are shared standards, data annotated against them can be understood and reasoned over by any system that implements the same ontological framework, regardless of organizational boundaries.

  • Formal semantics — the ontology carries machine-readable axioms that enable automated consistency checking, classification, and inference across datasets from different producers.

Ontology Support

The platform ships with support for defense-specific ontology extensions and provides an extensibility framework for mission-tailored vocabularies.

Defense Information Core Ontology (DICO)

DICO extends the BFO/CCO foundation with concepts specific to defense information management — intelligence reports, operational orders, mission plans, and associated metadata. It provides the vocabulary needed to represent and query defense-domain data in a semantically consistent way.

Joint Information Knowledge Ontology (JIKO)

JIKO adds joint interoperability concepts that span service branches and coalition partners. It defines shared semantics for entities, activities, and relationships that are common across joint operations.

Extensible Ontology Framework

Operators can extend the ontology with mission-specific vocabularies:

  • New classes and properties are defined using standard OWL/RDF syntax.

  • Extensions are validated against the BFO/CCO upper ontology to ensure consistency.

  • Deployed extensions are immediately available to the virtual knowledge graph and federated query engine.

  • Multiple extensions can coexist, enabling different missions to define domain-specific concepts without conflicting with one another.

OBDA Virtual Mappings

Ontology-Based Data Access (OBDA) creates virtual mappings from native data structures to ontology projections. These mappings are the bridge between the physical world of databases, streaming topics, and object stores and the semantic world of ontology classes and properties.

  • No data movement required — mappings are evaluated at query time. Data remains in its source system; no ETL pipelines, no materialized triples, no synchronization jobs.

  • Native data sources supported — relational databases, streaming topics, and S3-compatible object stores are all projected into RDF triples through their respective connectors.

  • Declarative mapping definitions — each mapping specifies the source table or topic, the target ontology class or property, and any column-to-property transformations. Mappings are defined once and updated only when the source schema changes.

  • SPARQL query interface — users query the virtualized ontology views using standard SPARQL. The OBDA engine translates SPARQL queries into optimized native queries against each source system.

Knowledge Graph Projection

The virtual knowledge graph (VKG) constructs semantic views over operational data, enabling graph-based reasoning and traversal without materializing a separate graph database.

RDF and SPARQL

The platform’s semantic layer is built on W3C standard RDF (Resource Description Framework) and SPARQL (SPARQL Protocol and RDF Query Language).

  • RDF provides the data model: every data element is represented as a subject-predicate-object triple.

  • SPARQL provides the query language: users write declarative graph queries that traverse relationships, filter by properties, and aggregate results.

Graph Traversal and Reasoning

The VKG supports multi-hop graph traversal, enabling queries that follow chains of relationships across entities and data sources. Reasoning capabilities allow the system to infer new relationships from existing data based on the ontology’s axioms — for example, inferring that an entity participating in a specific activity is also associated with the activity’s parent operation.

Natural-Language Query Interface

Non-technical users interact with the knowledge graph through a natural-language interface. Users describe their information needs in plain language, and the VKG translates the request into a structured SPARQL query against the ontology. This enables operators without query-language expertise to access the full depth of the semantic layer.

Next Steps