Use of CCO/BFO

[justinwb]

A group of participants in the data interoperability panel have proposed the use of Common Core Ontologies by the Solid Ecosystem.

They have shared an example use case in the form of a User Profile, and described its use of CCO as follows:

As an extension of CCO, the User Profile also inherits and re-uses Basic Formal Ontology (BFO) and OBO Relation Ontology. Use of the methodology and standards derived from CCO, BFO, and RO, provides an integration layer for an individual’s profile data that maintains a transparent and unified semantics across domains and applications.

Also related to the above were issues and pulls in the Data Interoperability Panel - solid/data-interoperability-panel#24, solid/data-interoperability-panel#14.

On October 21st, 2019 - Jacob McConomy, Jim Schoening, and Mark Jensen gave a detailed presentation on CCO / BFO. Minutes from that presentation where recorded, and Slides were made available.

Creating this issue for others in the community to consider the use of CCO/BFO/OBO within the Solid Ecosystem.

cc: @mark-jensen, @JKReynolds

[RubenVerborgh]

So "use" by itself is not a problem, in the sense that we all can use the ontologies we want.

The question is thus rather to what extent we want to push/recommend specific ontologies such as CCO, and how we go about that (regardless of the specific choice of ontology).

The current way of thinking (that I'm aware of) is that, in the long-term, we don't want to make such recommendations globally (with perhaps the exception of a small number of very common fields), but in the short- to mid-term, we might want to have a suggestion.

So we need to look at two things:

1. In general, what kind of endorsement/suggestion/… we want for ontologies?
2. Why is CCO/BFO a good choice for Solid, how does it compare to alternatives?

[kjetilk]

Generally, my thinking (which is probably influenced from community buzz) is that vocab consensus will emerge organically because of the network effect that successful usage will provide.

I suppose we need to be open to the possibility that this isn't really happening, and take schema.org as a counterexample, more steering might be needed. I think we need to have that debate on a principle level, as a debate on if and if so how we promote vocabs.

[TallTed]

The schema.org "ontology" (which term I use loosely, here) should also be a cautionary tale, as it is horrendously incomplete and even internally inconsistent -- significantly with incorrect range/domain statements, which can make working with data based on the schema.org "ontology" troublesome and even problematic.

I am not saying that similar issues pertain to CCO/BFO/OBO; only, that endorsing any external ontology, no matter its popularity or cachet, has the potential to open a can (or barrel) of worms that may be better left on the shelf.

[csarven]

Are there (experimental) Solid applications making use of those ontologies?