Solid environmental material as a subclass of ice

[Clarissa-Feio]

Hello,

I just wanted to let you know that you have solid environmental material as a subclass of ice.

Kind regards, Clarissa.

[pbuttigieg]

Hi @Clarissa-Feio thanks!

Oddly enough, this isn't an error, but a challenge in handling the astrogeological definition of what an "ice" is. There are planetary conditions where you can have rather hot ice, kept solid by gravitational forces.

For vanilla Earth ice, please use "water ice".

This is a good reason to provide annotators with subsets so they don't have to delve too deeply into weird planetary science semantics for Tellurian concerns.

[smrgeoinfo]

is there any other child of ice besides 'solid environmental material'? If not it seems like it would be much more sensible to have 'astrogeologic ice' as an alternate label for 'solid environmental material'. Using ice for the label at the top of the hierarchy shown is not at all consistent with any usage I'm familiar with, and will cause confusion and laughing-- charcoal is a kind of ice??? dust is a kind of ice-- that doesn't make sense under any definition I've ever heard!

I suggested in another thread around here somewhere that ice in normal usage means 'a solid that is normally a fluid (gas or liquid) under standard environmental conditions (STP)'.

[pbuttigieg]

@smrgeoinfo

Using ice for the label at the top of the hierarchy shown is not at all consistent with any usage I'm familiar with, and will cause confusion and laughing-- charcoal is a kind of ice??? dust is a kind of ice-- that doesn't make sense under any definition I've ever heard!

Yes, agreed - I'm trying to find a way to handle this sensibly the issue is related to the next quote...

'astrogeologic ice' as an alternate label for 'solid environmental material'.

Astrogeological ice is another class defined by melting point rather than actual state.

At any rate, I think we should shelf and remove the top level "ice" from ENVO for now, until we get more expertise on this in the editorial team and have a more defined need for it.

[rduerr]

Interesting discussion! And it points out the need to identify the domain(s) a term is relevant in, and the ability to disambiguate meaning based on that!

Agreed about the suggested solution!

[cmungall]

Folks, this is clearly an error! There is no discussion to be had here!

[cmungall]

At any rate, I think we should shelf and remove the top level "ice" from ENVO for now, until we get more expertise on this in the editorial team and have a more defined need for it.

We would need a home for the asserted children:

'ammonia ice' 'blue ice' 'carbon dioxide ice' 'carbon monoxide ice' 'methane ice' 'water ice'

There is a common-sense short term solution: remove the logical definition:

'environmental material'
 and ('has quality' some 
    ('quality of a solid' or frozen))

this is clearly too inclusive, and leads the reasoner to classify 'dust', 'charcoal' etc under it

There may be other things we want to iterate on, but let's not get stuck in a rabbit hole, perfection is the enemy of the good.

I will go ahead and make a PR

@smrgeoinfo

Using ice for the label at the top of the hierarchy shown is not at all consistent with any usage I'm familiar with, and will cause confusion and laughing-- charcoal is a kind of ice??? dust is a kind of ice-- that doesn't make sense under any definition I've ever heard!

quite! except I am crying not laughing

my PR will remove these egregious classifications

[cmungall]

Closing this, as it was the same issue as #782

Current view in OLS:

There are some separate issues here for which I will make a new ticket

[dosumis]

There is a common-sense short term solution: remove the logical definition:

'environmental material'
 and ('has quality' some 
    ('quality of a solid' or frozen))

This is a great example of over-use of Equivalent Class Definitions. If you are happy that all of these clauses are true, make subClassOf restrictions. To put it another way, I'm sure you're happy to say that all ice is solid or frozen. I'm pretty sure you don't mean to say that all solid environmental material is ice, but that's what you've done here!

Coming up with good Equivalent Class definitions is hard. Don't add one unless you are confident it will work.

[kaiiam]

Thanks for the explanation @dosumis and @cmungall who in the associated PR said:

First-order logic (and by extension OWL) is a very limited language for expressing definitions for basic scientific concepts. OWL excels when it comes to definitions for compositional concepts. "methane ice" may be such a concept (ice and composed-primarily-of some methane)

Makes sense that equivalence classes are best for compositional classes X ice rather than the basic class ice

@cmungall did you want to intend this issue? it's still open ATM.

[kaiiam]

I guess in the above axiom:

'environmental material'
 and ('has quality' some 
    ('quality of a solid' or frozen))

it's a logical OR in the 'quality of a solid' or frozen hence the inclusion of terms like charcoal? This makes sense as charcoal has the axiom 'has quality' some 'quality of a solid'

[cmungall]

I guess in the above axiom

the above isn't an axiom - it's a class expression

whether the class expression is used in a subClassOf axiom or equivalence axiom has a bearing on the answer to your question

(my PR changed it from EC to SC)

it's a logical OR in the 'quality of a solid' or frozen hence the inclusion of terms like charcoal? This makes sense as charcoal has the axiom 'has quality' some 'quality of a solid'

Yes, if the expression is used in an EC then you will see charcoal inferred to be a subclass of ice, which is clearly wrong

[cmungall]

did you want to intend this issue? it's still open ATM.

This can be closed as my PR fixed it

[kaiiam]

Thanks @cmungall for clarifying the terminology for I had been saying axiom as if it meant an EC or SC axiom with the following class expression. I guess it would be best to be better for me to discuss it that way, rather than saying just an axiom?