monarch-initiative / vertebrate-breed-ontology

https://monarch-initiative.github.io/vertebrate-breed-ontology/
8 stars 0 forks source link

Fresians are independent continuants but Holsteins are not #53

Open cmungall opened 1 year ago

cmungall commented 1 year ago
image

Why is a Friesian an independent continuant, yet a Holstein is not (or is at least ontologically uncommitted?)

Maybe Fresians have been mixed up with Fregeians?

Are Holsteins actually immaterial entities? Is it because Bos torus being a manifold is a BFO 2D entity? In that case can Fresians really inherit from Bos taurus?

But more seriously, I suspect that there is something up with the subclass having some domain constraint that induces the subclasses, which is materialized by robot in the release pipeline

(the FMA issue is mentioned separately #52)

sabrinatoro commented 1 year ago

Friesian being an independent continuant appears to be inferred based on the "located in" value [country] subclass Screen Shot 2022-09-23 at 6 59 35 PM

This inference is made for all breeds that have a subclass including country. The parent term does not have any relation to country, and therefore does not get the inference to "independent continuant".

What does it mean in terms of the ontology? I don't know. Is having the country of existence as a subclass the wrong way to do it? Is "located in" the wrong relation to use?

cmungall commented 1 year ago

A cob import should fix this, as it will asserted the root of ncbi taxon to be an IC, rendering the fresian inference correct but redundant

cthoyt commented 1 year ago

is there a joke in here about spherical cows?