Image 1 of 1
Image 1 of 1
Image 1 of 1
Image 1 of 1
A one-day, hands-on course in SHACL — the W3C standard for validating RDF data. A knowledge graph tells you what the world looks like; SHACL is how you say what it is allowed to look like, and get a machine to check. OWL, being open-world, will not reject a bad record for you. SHACL will.
You spend the day writing shapes by hand and reading the validation reports they produce. That is the skill this course exists to give you. The engine, the tooling and the SPARQL escape hatch are covered, but late and deliberately — a validator you can only click is not a skill.
By the end of the day you will be able to:
Explain why SHACL exists — which data-quality problems it solves that OWL, by design, does not
Write shapes from scratch — declare targets and property paths in Turtle, without tooling doing it for you
Express real rules — apply the core constraint components to the data-quality conditions your own graph actually needs
Read a validation report — trace every violation back to the shape that produced it, and know what to do about it
Know when to escape to SHACL-SPARQL — and, more usefully, when not to
Run an engine over your own data — and act on what comes back
Practitioners who already work with RDF and now need their data to hold a standard: data engineers, ontologists, knowledge-graph developers, and data architects responsible for quality.
You should already be comfortable with: reading and writing Turtle, and the basics of classes, properties, and the TBox/ABox distinction. This is an advanced module — it does not start from "what is a triple".
Six hours, 09:00–15:00, lunch and two breaks included. Lecture and practice alternate all day; you are never more than an hour from writing something yourself.
You get access to a browser-based playground — no local install, nothing to configure. Every exercise ships with data that fails validation on purpose, so you practise diagnosing broken data rather than admiring clean data. The playground stays open after the session for your own experimenting, and the day closes with a quiz.
There are deliberately more exercises than time. The set is not meant to be finished; the first three carry the core skill and the rest are there for you to take home.
Shapes you wrote yourself, and the confidence to write more
The ability to read a validation report as a diagnosis rather than a wall of text
Continued access to the playground and the full exercise set
A clear sense of where SHACL's core components end and SPARQL begins
A one-day, hands-on course in SHACL — the W3C standard for validating RDF data. A knowledge graph tells you what the world looks like; SHACL is how you say what it is allowed to look like, and get a machine to check. OWL, being open-world, will not reject a bad record for you. SHACL will.
You spend the day writing shapes by hand and reading the validation reports they produce. That is the skill this course exists to give you. The engine, the tooling and the SPARQL escape hatch are covered, but late and deliberately — a validator you can only click is not a skill.
By the end of the day you will be able to:
Explain why SHACL exists — which data-quality problems it solves that OWL, by design, does not
Write shapes from scratch — declare targets and property paths in Turtle, without tooling doing it for you
Express real rules — apply the core constraint components to the data-quality conditions your own graph actually needs
Read a validation report — trace every violation back to the shape that produced it, and know what to do about it
Know when to escape to SHACL-SPARQL — and, more usefully, when not to
Run an engine over your own data — and act on what comes back
Practitioners who already work with RDF and now need their data to hold a standard: data engineers, ontologists, knowledge-graph developers, and data architects responsible for quality.
You should already be comfortable with: reading and writing Turtle, and the basics of classes, properties, and the TBox/ABox distinction. This is an advanced module — it does not start from "what is a triple".
Six hours, 09:00–15:00, lunch and two breaks included. Lecture and practice alternate all day; you are never more than an hour from writing something yourself.
You get access to a browser-based playground — no local install, nothing to configure. Every exercise ships with data that fails validation on purpose, so you practise diagnosing broken data rather than admiring clean data. The playground stays open after the session for your own experimenting, and the day closes with a quiz.
There are deliberately more exercises than time. The set is not meant to be finished; the first three carry the core skill and the rest are there for you to take home.
Shapes you wrote yourself, and the confidence to write more
The ability to read a validation report as a diagnosis rather than a wall of text
Continued access to the playground and the full exercise set
A clear sense of where SHACL's core components end and SPARQL begins