search

OBOFoundry / OBOFoundry.github.io

Metadata and website for the Open Bio Ontologies Foundry Ontology Registry
http://obofoundry.org
Other
164 stars 201 forks source link

Adding numerics to axiom #849

Open lschriml opened 5 years ago

lschriml commented 5 years ago

@pbuttigieg requested discussion of adding numerics to axioms.

Here's an issue that resurfaces frequently in ENVO and SDGIO development, which I think would need an OBO-wide policy:

Many users use classification systems with hard boundaries based on some sort of measured value. For example, a climate zone with an upper temperature limit of 35*C.

Traditionally, we've shied away from axiomatising these sort of thresholds, going instead for more qualitative differentia. However, in some cases, the fiat/numeric boundary is really the only difference.

I'm at a stage in ENVO/SDGIO development where I can't really keep saying "we don't do that" (important partners like UN Environment would find this very odd).

Can we create a policy/guide on how this should be approached? I assume the axioms will shortcut a combination of PATO, IAO, UO semantics.

cmungall commented 5 years ago

@pbuttigieg I like how you can go to the SIO website and see their main DPs nicely documented with examples, e.g https://github.com/MaastrichtU-IDS/semanticscience/wiki/DP-Measurements

I would like a simple model like this for OBO.

See also https://github.com/pato-ontology/pato/issues/101

jamesaoverton commented 5 years ago

There's a pressing need to document this stuff for OBO. Unless somebody (anybody!) else wants to do it, I will volunteer.

cmungall commented 5 years ago

Any progress on this @jamesaoverton (on tomorrow's agenda)

ramonawalls commented 5 years ago

PPO has created new data properties called upper count, lower count, upper percent, and lower percent. We use these in GCAs to define a set of parent classes for the "presence" of structures. Each presence class can have a present or absent subclass. The subclasses are further defined by a GCA that sets the limits of those upper and lower values. For example, if a structure is absent, either the upper count or upper percent is zero. If a structure is present, the lower count or lower percent is >0.

pbuttigieg commented 5 years ago

Thanks all. Our use cases would be, e.g., thresholds for canopy density, ice packing, and GDP. I realise this drags units into the mix, but if we have a good strategy to coordinate our core this may be tractable.

nlharris commented 2 years ago

is this still relevant?

matentzn commented 2 years ago

I read this issue 10 times now not sure what it means, but my guess is that this is a request to add a page to OBO somewhere that explains:

1) how to best document data properties 2) how to best document design patterns in an ontologies related to the modelling of quantities

This is probably generally still a good idea.

ramonawalls commented 2 years ago

This is still relevant and important for C-Path's work as well as the quantitative trait work I have been involved with. @jamesaoverton @matentzn @beckyjackson is there bandwidth to make this part of our current contract and maybe include it in OBOok?

jamesaoverton commented 2 years ago

@ramonawalls Yes, our plan is to use some of the hours on our contract with you to finish up the quantitative values work that I've shared at COB and OBI workshops, and to integrate that with OBO documentation. I expect to start that in the next few weeks.

PS: That funding to support this work is greatly appreciated!

cmungall commented 1 year ago

On Mon, May 23, 2022 at 10:08 PM Nico Matentzoglu @.***> wrote:

I read this issue 10 times now not sure what it means, but my guess is that this is a request to add a page to OBO somewhere that explains:

  1. how to best document data properties

See also: https://github.com/OBOFoundry/COB/issues/195

  1. how to best document design patterns in an ontologies related to the modelling of quantities

See also: https://github.com/OBOFoundry/COB/issues/35

But Pier's question also involves axiomatizing TBoxes using quantities/measurements, which is not straightforward. Or rather: it is not straightforward determining precisely what the actual statement we want to axiomatize is

See: https://github.com/oborel/obo-relations/issues/155

This is probably generally still a good idea.

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/OBOFoundry.github.io/issues/849#issuecomment-1135414298, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAMMOIWIJWO47FOBNVT47LVLRP4NANCNFSM4HBPXBFQ . You are receiving this because you commented.Message ID: @.***>