Expand & Integrate Cardinality checking

[cuddihyge]

• [ ] Kevin. how do we leverage [ASSIST] (https://github.com/ge-high-assurance/RACK/wiki/RACK-Assist-Bin-Tools)
• [ ] Abha
• [ ] sadl rules examples seem too difficult to maintain

Goal is to expand restriction-checking:

• to must be objectOf at least 1
• unique?
• re-check qualified cardinality?
• abstract class like REQUIREMENT ?

Targets:

• the simplicity of SADL cardinality : domain experts not semantic experts.
• simple to run : plug into SPARQLgraph restriction-checker
• stretch: the power of SHACL
• clear documentation using OWL terminology of what is checked, including retro documenting ingestion in these terms (e.g. qualified cardinality): build on [data-verification wiki] (https://github.com/ge-high-assurance/RACK/wiki/Data-Verification)

For the model, does SADL/OWL have these (I see inverseOf). Ideally we want a clean way to model the constraints and a clean way to run them (currently we run standard cardinality in SemTK with Java building SPARQL).

Don't want to use inverseOf and add it to the model in a way that it might be instantiated and now all the queries need to be complex or need an inference engine.

[cuddihyge]

Hope to have a POC of SHACL for the release. If it is successful we can use the Sep-Mar time to improve the tooling

[weisenje]

We addressed this Issue with our integration of SHACL into SemTK/RACK. @kityansiu feel free to close it