Re: SHACL shapes generation

Hi Kashif,

You may also be interested in Dragon [1], which will transform almost any
of Uber's schemas into a  SHACL shapes graph. A couple of weeks ago, I gave
a presentation [2] in which I converted (a representative, demo-safe sample
of) Uber's metadata catalog [3] to RDF. Workflows like this involve a
schema-level mapping (YAML to SHACL RDF in this case) and a data-level
mapping (JSON to instance RDF) in parallel. By design, the two mappings are
consistent, such that a dataset which conforms to the original schema maps
to an RDF graph that conforms to the generated SHACL shapes graph. Dragon's
schema language can be seen as a proper subset of SHACL.

Best regards,

Josh


[1] https://eng.uber.com/dragon-schema-integration-at-uber-scale/
[2] https://www.meetup.com/Category-Theory/events/277185694/
[3] https://eng.uber.com/metadata-insights-databook/


On Mon, Apr 19, 2021 at 5:55 AM Sven Lieber (UGent-imec) <
Sven.Lieber@ugent.be> wrote:

> Hi Kashif,
>
>
>
> We investigated the creation of data shapes from different angles which
> both might be relevant for you as they concern the creation and use of data
> shapes.
>
> On the one hand we looked into which constraint types are being used [1,
> 2],
>
> and on the other hand we investigated the creation of data shapes by users
> using a visual language [3, 4, 5] for which we also developed a tool [6].
>
>
>
> Regarding automatic generation of data shapes from ontologies I am also
> aware of Astrea [7] to which we aligned in our work [1].
>
> These works (and related work mentioned in the papers) may give you some
> hints for further reading.
>
>
>
> Best,
>
> Sven
>
>
>
> [1] http://ceur-ws.org/Vol-2721/paper584.pdf
>
> [2] https://zenodo.org/record/4154456
>
> [3]
> http://www.semantic-web-journal.net/content/visual-notations-viewing-and-editing-rdf-constraints-unshacled
> (under review)
>
> [4] https://w3id.org/imec/unshacled/spec/shape-uml
>
> [5] https://w3id.org/imec/unshacled/spec/shape-vowl
>
> [6] https://github.com/KNowledgeOnWebScale/unshacled
>
> [7] https://link.springer.com/chapter/10.1007/978-3-030-49461-2_29
>
>
>
> Kind regards,
>
>
>
> *SVEN LIEBER*
> PhD researcher semantic intelligence
> IDLab <http://idlab.technology/>,  imec <https://imec.be/> research group
> at Ghent University <https://www.ugent.be/en>
> AA-tower | Technologiepark Zwijnaarde 122 | 9052 Ghent, Belgium
>
> t: +32 9 331 49 56
>
> [image: iconmonstr-globe-5-24] <https://sven-lieber.org/>[image:
> iconmonstr-twitter-5-24] <https://twitter.com/SvenLieber>[image:
> iconmonstr-linkedin-5-24]
> <https://www.linkedin.com/in/sven-lieber-1b2b70131/> [image:
> iconmonstr-key-11-24] <https://sven-lieber.org/pgp-key>
>
>
>
> *From:* Kashif Rabbani <ch.kashirabbani@gmail.com>
> *Sent:* Monday, April 19, 2021 10:20 AM
> *To:* semantic-web@w3.org
> *Subject:* SHACL shapes generation
>
>
>
> Hi everyone,
>
>
>
> We are studying SHACL shapes generation. I was wondering which tools or
> approaches are being used by the community to generate SHACL or ShEx shapes
> for large RDF graphs?
>
>
>
> Regards,
>
> Kashif Rabbani.
>
>
>