Comments on the diagram panel

• Bubbles overlap in the diagram – hard to see everything
• Why is the panel called “Data model ..."?
  • the ‘data’ shown are really semantic metadata, are they not?
  • also, ‘model’ suggests a more complete set of information, and what is shown are only immediate semantic properties.
  • Perhaps “Semantic graph view for …” might be more appropriate than “Data model for …” ?
  • After double-clicking on a class or property in the graph view, the entire web page is redisplayed and left positioned at the top, with the home page banner, etc. One has to scroll down in order to re-engage with the graph.
  • Admittedly this action causes new data to be loaded and a refresh is required, but perhaps the graphical viewer should be displayed in a panel that is independent of the main page so that navigating the graph can be more continuous and without being visually reset out of view?

TO DO

• ...

DONE

• ...

Too much clutter at the top. Causes too much scrolling.

The user-interface still looks a bit old-school, like the whole EDMC site. If there's budget now I would suggest to ask a designer to give it an upgrade.

It is not a really modern site.

Instances should be treated separately.

• We actually reserved the hostname data.edmcouncil.org for that where each individual could get its own "Knowledge Graph IRI" (KGIRI) in the form https://data.edmcouncil.org/resource/<uuid> where UUID is a UUID type 5 hash of some keys.
• You don't want any "meaning" in the instance IRI because one instance can have many types and be used as a test object for many different user story tests.

• If you look at this page you simply see too many individuals: https://spec.edmcouncil.org/fibo/ontology/FBC/FunctionalEntities/MarketsIndividuals/

In many cases, I agree, but not all instances are alike!

We need to separate:

• Enumerated values (since we’re using things not strings) which are part of the ontology and should not be “treated separately”
• Examples, for illustrative purposes
• Test data, which will often be artificial to stretch one or more (ideally more) ontologies with minimal volume
• Reference data (an example of which is the MarketsIndividuals) Jacobus referenced where we’re transforming external data and publishing it in a FIBO-compatible RDF form with our own URIs

Also, without obvious links to the other products, especially the OWL, it looks like another glossary.

What's missing:

• Link to the OWL file (in multiple formats)
• Link to the corresponding git versions
• List of all the versions including their version IRIs (https://spec.edmcouncil.org/fibo/ontology/<branch>/<tag>/....)
• List of all the git commits including their associated JIRA tickets
• List of all the pull requests including test results and logs
• All shown file names like "Corporations.rdf" should be links
• All shown URLs / IRIs should be clickable and resolvable

Especially missing is an overall graph.

Missing: Cory's idea was to ask NL questions of intent and then navigate from there.

Missing: Dennis's Lego Blocks idea - caned models for individual purposes.

And it ignores the requirements we spent many weeks gathering and documenting in our own Wiki with, and for, a variety of stakeholders.

There were 5 years of requirements collected and seemingly all ignored.

Generic Content

(applies to any type of Element)

Note: any URLs (and ontology URIs) should be recognized and made clickable, including references from annotations.

Glossary/Business User View

Always shown

• Primary label
• Synonyms
• Definition
• Explanatory Note
• Logical Description (generated)
• Type (i.e. Class/Noun - open issue what terminology)
• Ontology/Module/Domain hierarchy
• Maturity

TO DO

• ...

DONE

• ...

On request

• Term origin
• Adapted from
• Definition origin
• (not Editorial Note which is for maintainers)
• IRI
• Version

Taxonomy and Structure Views

All of the above

• Access path (different groupings, probably hierarchic, and navigations reflecting different usage context)
  
  

Ontologist View

Always shown

All of the above, using the actual predicates, except:

• Logical Description (generated)

On request

• All annotations