01
Commerce Language & Ontologies
Controlled concepts for principals, products, capabilities, state, authority, evidence, actions and outcomes.
Research Area 12
Data, Semantics &
Interoperability
Researching the language, data and protocol fabric that lets companies, countries, agents, machines and products understand one another.
From data silos to a connected commerce ecosystem
Making commercial meaning work together across the world.
Commercial systems often use different classifications, schemas, units, definitions and identifiers. Those mismatches create hidden economic cost and make automation brittle.
Interoperability can reduce repeated translation and reconciliation while preserving local and sector-specific meaning.
See the interoperability stackConnected global data
Linking commercial and regulatory evidence across borders.
Shared meanings
Common concepts without erasing local context.
Interoperable systems
Protocols that allow diverse systems to exchange meaning.
The interoperability stack
From raw data to shared understanding.
We study the layers required to turn diverse real-world commerce data into a trusted, interoperable foundation for analysis and action.
Data sources
Companies & suppliers
Products & catalogs
Transactions & payments
Logistics & transport
Regulations & policy
Government data
Identity & Linking LayerEntity resolution and graph links
Semantic LayerOntologies, schemas and common definitions
Knowledge LayerRules, context and relationships
Protocol LayerDiscovery, requests, offers, authority and evidence
Trust & Governance LayerQuality, provenance, contradiction and access control
Data users & applications
Companies & platforms
AI agents & analytics
Governments & regulators
Country-native systems
Search & retrieval
Global decision systems
Five public research programs
Data, Semantics & Interoperability Research Programs
02
Commerce Protocols & Standards
Interoperable patterns for discovery, requests, offers, negotiation, authority, evidence, fulfillment and settlement.
03
Search & Retrieval Science
Retrieval across structured data, documents, current state, rules and evidence.
04
Economic Knowledge Infrastructure
Provenance-aware knowledge that preserves version, source, authority and contradiction.
05
Semantic Friction & Translation
Measurement and reduction of losses created by language, units, classifications and schema mismatch.
Semantic friction laboratory
Meaning mismatches create economic cost.
We study where information loses fidelity across language, units, classifications, identifiers and schemas—and whether those losses can be measured as economic friction.
Explore our methodsLANGUAGE
Same product, different commercial terms.
UNITS
Dimensions and quantities translated inconsistently.
CLASSIFICATION
Different product or regulatory taxonomies.
SCHEMA
Fields align syntactically but not semantically.
Search & retrieval science
Retrieval should span more than documents.
Commerce questions may require structured data, documents, current state, rules and evidence at the same time. Research should preserve authority and provenance across those sources.
Structured data
Documents
Current state
Rules
Evidence
Economic knowledge infrastructure
Provenance and contradiction should be first-class data.
Provenance-aware knowledge
SourceWhere did the claim originate?
AuthorityWho is entitled to assert it?
VersionWhich definition or rule was in force?
TimeWhen was the claim true?
Contradiction handling
ConflictTwo sources disagree.
ContextDifferent jurisdictions or scopes apply.
UncertaintyEvidence is incomplete or stale.
ResolutionEscalate, compare authority, or preserve disagreement.
Key research questions
Questions we are exploring.
What commercial meaning must remain stable across systems?
How should ontologies evolve without breaking existing integrations?
Can semantic friction be measured as an economic cost?
Our research boundary
Open standards where ecosystem value is high.
Public standards and protocols may be released where ecosystem value is high. Internal data models, proprietary mappings and protected knowledge assets may remain private.