Scope of IAO - independent variable

[Bjoern Peters]

We had some discussions on the scope of IAO at the workshop, and I
thought there was a majority (not: consensus) that wanted to keep IAO
small and let OBI take all 'investigation specific' terms. 'independent
variable' is clearly investigation specific, and I was hoping to have it
in the 'plan and planned process branch', but it is under IAO now.

Can we move it? If not, it seems that other terms like 'study design',
'investigation' etc. should also be moved to IAO. I would be very
opposed to that.

- Bjoern

[Alan Ruttenberg]

My preference would be to start organizing, even if it were maintained
within OBI, an upper investigation ontology that is still smallish
that contains such terms. It should contain only terms that
researchers in the other natural sciences could take and use without
adaptation.

We might accomplish this technically by use of annotations/scripting.

However this work is broader than OBI and time spent on it should be
easily beneficial to a wider audience.

-Alan

[Barry Smith]

At 02:48 PM 2/17/2009, Alan Ruttenberg wrote:
>My preference would be to start organizing, even if it were maintained
>within OBI, an upper investigation ontology that is still smallish
>that contains such terms. It should contain only terms that
>researchers in the other natural sciences could take and use without
>adaptation.

Good idea.

>We might accomplish this technically by use of annotations/scripting.

Less good idea.

>However this work is broader than OBI and time spent on it should be
>easily beneficial to a wider audience.

Yes.
BS

[Bjoern Peters]

Flagging some terms in OBI for potential future generalization into an
investigation ontology is fine with me. Everything that would
immediately result in an id space and responsibility schism like IAO/OBI
would have to be approved at a OBI coordinator call. I would oppose it,
as I see practical downsides: additional difficulty in identifying who
is responsible for which term, and difficulty in coordinating any
changes. If you doubt that this is a problem, look at the GDC branch in
OBI as it is now.

Also: where would we stop? Surely terms like organization, molecule,
acquisition are not OBI specific, but neither are they investigation
specific. I doubt that much more than 20% of OBI terms can only be used
in biomedical investigations. Do we split those in a legal ontology,
object ontology and human interaction ontology?

- Bjoern

[James Malone]

I can see the sense in your vision Alan and in theory I totally agree, but I also worry about the impact on OBI development to begin to attempt another divergence in strategy now. Our efforts are already diverted enough with IAO and BFO. I would hope that our strategy for building OBI is such that the higher level could be extracted anyway in the manner you mention Alan without having to reorganise things, but to be honest this is not my major concern right now.  The scope if biomedical and that is my focus.  I would encourage this to be opened up for wider discussion Alan at an OBI coord meeting/mail if it is something you'd like to propose (even as just an idea).  What I would be really against is reorganising OBI by stealth.

James