Intermediate · About 30 minutes with practice

Make relationships precise enough to use

Build a small knowledge graph, then define what its entities and relationships actually mean.

Make this yours

A six-entity graph, a compact vocabulary and a provenance query.

Start with your own answer ↓
Speculative Academy artwork: the seed ark opens.
Imagined world · The seed ark opens

A moment before the explanation

How do a graph, an ontology and a validation rule do different jobs?

Give yourself a first answer, even if it is incomplete. Keep it in mind as you read—or write it in your field note.

01

Start with a question, not a diagram

Ask one question your graph should answer: which source supports this claim, and which artifact depends on it? Name stable identifiers for source, claim, artifact, reviewer, task and review. A line between two boxes is useful only when the relationship has a precise meaning.

RDF represents statements as subject–predicate–object triples. Reusing identifiers lets different statements connect. That is a data model, not a guarantee that the statements are true. Evidence and domain judgment still matter. Keep a small example you can understand before adding graph infrastructure.

02

An ontology gives terms a shared meaning

Define each entity type and relationship with examples and non-examples. “Created by,” “reviewed by” and “approved by” should not collapse into a vague “associated with.” A review may find defects; an approval may authorize a particular action. Their effects differ.

OWL can express logical relationships among classes and properties. An application may also need explicit validation rules and database constraints. For example, require a source reference before accepting a claim into a digest. Do not assume that absence of a fact automatically means the fact is false; choose semantics deliberately.

03

Carry evidence and time through the graph

Attach provenance to the statement being made, not just to the whole collection. A claim may be supported by a dated passage and later superseded. Distinguish when something happened from when your system observed it. Those times answer different questions.

Test a query against a contradictory example and a missing edge. Can a reader trace the result back to a source? Can the graph distinguish a draft from an approved artifact? A useful graph makes a consequential question easier to answer correctly. A beautiful visualization is optional.

Inspect a small knowledge graph

Follow the claim back to evidence.

Claim · release-date

The draft states a release date. Its identifier stays stable so evidence and revisions can refer to the same claim.

Supported by
source:announcement-a
Included in
artifact:digest-v1
Confidence
Requires source inspection

The arrows are a simplified reading path. The records expose the actual relationship names.

A worked example · Illustrative

A small evidence vocabulary

source:announcement-a → supports → claim:release-date
claim:release-date → included-in → artifact:digest-v1
review:42 → evaluates → artifact:digest-v1
review:42 → found → issue:missing-timezone
artifact:digest-v2 → revises → artifact:digest-v1

Query: Which artifacts depend on a disputed source?
Constraint: reviewed-by does not imply approved-by

Put the idea to work

Your practice

  1. Choose a question and create six entities with stable identifiers.
  2. Define four relationship types with examples and counterexamples.
  3. Add provenance and time to one disputed statement.
  4. Write a plain-language query and manually verify its expected answer.

Make the idea available when you need it

Your field note

Keep your first answer, make an artifact and try the idea somewhere new. Save only what you want stored on this device.

Check one distinction

An artifact has a “reviewed by” edge. May the system infer approval to publish?

Take this practice into your own agent.

Copy a coaching brief, then choose your host in the Academy installer.

Choose a host ↗

Your text stays in this page until you save or export. Local saving can be unavailable in restricted browsers. Curriculum 2026-09-15.1. Import, export and browse your notebook ↗

Follow the evidence further

Primary sources & deeper study

Selected 15 September 2026. Provider docs govern current setup; older courses and engineering articles provide foundations.

W3C · Standards primer

RDF 1.1 Primer

Subjects, predicates, objects, identifiers and named graphs.

W3C · Standards primer

OWL 2 Primer

Classes, properties and logical relationships; pair with validation and domain review.

Starlight Intelligence System · Versioned source

Operational work graph

Event identity, proof gates, provenance and memory promotion. Read as an architecture contract.

Take the method into your work

Map an actual workflow
Another question to follow

Make benevolence a design decision