EKG Principles
The Enterprise Knowledge Graph Forum publishes ten EKG Principles as guidelines for developing and operating an Enterprise Knowledge Graph. Together they cover stable identity, machine-readable meaning, distributed access, contextual truth, self-description, measurable quality, business purpose, policy control, heterogeneous ecosystems, and open standards.
Buyer question
“How closely does Legra support the ten EKG Principles, and which parts still depend on our own models, policies, infrastructure, and operating practice?”
Legra fit
| EKG Principle | Legra fit | Status |
|---|---|---|
| 1. Identity | Legra preserves stable IRIs for RDF resources, workspaces, named graphs, and semantic artifacts. The organization publishing an identifier remains responsible for its permanence and public web resolution. | Partial fit |
| 2. Meaning | RDF, RDFS, OWL, SHACL, shared vocabularies, and reasoning make definitions machine-readable and directly linkable from data. A deployment must still model and govern the meaning of its own data points. | Native fit |
| 3. Distributed | Encrypted workspaces can be hosted and replicated across peer nodes while gateways and workspace routing provide access without making one central database the authority. | Native fit |
| 4. Open World | RDF’s open-world model, named graphs, branches, provenance, and source-specific assertions let multiple contextual versions coexist. Domain-specific reconciliation rules remain an application concern. | Native fit |
| 5. Self-describing | Graph metadata, shapes, provenance, catalogs, policies, DPROD, and DCAT descriptions can live beside the data they describe and follow the same versioning and governance lifecycle. | Native fit |
| 6. Measurement | SHACL validation reports and graph-based quality metadata provide measurable evidence, but each organization must define fitness criteria, thresholds, monitoring, and action ownership. | Partial fit |
| 7. Business Orientation | Use Case Trees and the Composable Business surface connect semantic artifacts, Stories, personas, outcomes, and data products to an explicit business purpose. | Native fit |
| 8. Control | Workspace membership, encryption, branch controls, validation gates, signed history, and provenance enforce important policies. Fully modeled and automatically enforced data-point-level business, entitlement, and privacy policies remain broader than the current product surface. | Partial fit |
| 9. Ecosystem | RDF interoperability, multiple query languages, loaders, ETL, Apps, Data Products, and peer nodes support heterogeneous sources and technologies. Source-specific service-level agreements remain an operational responsibility. | Partial fit |
| 10. Standards | Legra is built around RDF, SPARQL, SHACL, OWL, IRIs, Linked Data, and related open semantic standards rather than a proprietary graph model. | Native fit |
What Legra can say
Legra provides an architectural substrate for implementing the EKG Principles rather than treating them as a checklist stored outside the graph:
- Identities, definitions, policies, quality rules, provenance, and business purpose can all be represented as linked, queryable graph data.
- Branches and commits govern changes to data and the semantic artifacts that explain or constrain it.
- Distributed, encrypted workspaces avoid making one database or platform the sole authority for enterprise knowledge.
- Validation, reasoning, events, and provenance turn semantic models into operational behavior and evidence.
- Use Case Trees connect reusable semantic and executable components to the business outcomes that justify them.
Assessment boundary
The principles describe properties of an enterprise knowledge ecosystem, not only database features. Legra cannot make arbitrary source data meaningful, assign enduring public identifiers on an organization’s behalf, define useful quality thresholds, supply business justification, or establish operational service levels automatically.
An assessment should therefore distinguish the Legra capabilities in the table from the models, policies, publication infrastructure, and governance practice supplied by the organization operating the EKG.