@janmartinkeil XProc is used to define declarative XML data pipelines, involving pipeline steps like loading data, validating it against a schema, transforming it via XSLT. I would like to have a way for RDF to define declarative pipelines for loading RDF via , validating it against e.g. SHACL, and transforming it with SPARQL. I hope that this clarifies my question, let me know if you have further pointers - and thanks for the existing ones. I found this implementation https://github.com/rdf-pipeline/framework which is old and not maintained, but the concept is what I have in mind.
@fsasaki One could build that based on PROV-O, as I similarly did in #ABRCTO (https://github.com/fusion-jena/abecto). But I'm not aware of any widely used/accepted vocabulary that enables declarative descriptions of #RDF processing/pipelines. #SemanticWeb
@janmartinkeil that looks cool, thanks for the pointer! I will have a look.
@fsasaki I just realized, that I should have linked P-Plan (http://vocab.linkeddata.es/p-plan/index.html), not PROV-O (https://www.w3.org/TR/prov-o/), above.
@fsasaki we started building something (early stage) using P-Plan, DCAT and LDES for semantically describing streams:
* Spec: https://w3id.org/sds/specification
* Implementation of some runners: https://github.com/ajuvercr/nautirust
@pietercolpaert thanks a lot, Pieter! I will look into the implementation.
@fsasaki https://www.bobdc.com/blog/pipelining-sparql-queries-in-m/ describes one approach and lists a few others.
@fsasaki Not sure, what exactly you are looking for. If it is a way to represent workflows/pipelines, have a look into PROV-O (https://www.w3.org/TR/prov-o/) and related vocabularies. If it is a tool, have a look on ROBOT (http://robot.obolibrary.org/).