domain and range of 'instant before' and 'instant after' seem too permissive

[avsculley]

A temporal instant t2 (a instance of a zero-dimensional temporal region) is after another temporal instant t1 iff there exists some instance of a temporal interval ti1 such that t1 is the starting instant of ti1 and t2 is the ending instant of ti1.

As is, this definition, and the domain-range restrictions of 'zero-dimensional temporal region,' seem to allow for a collection of separated temporal instants (a zero-dimensional temporal region that is not a temporal instant) to be either the starting instant or ending instant of a temporal interval. This can't be true.

I suggest that the parenthetical be removed from the definition and the domain-range restrictions be changed to 'temporal instant.'

The same reasoning holds for 'instant before.'

Edit: If you agree, I can make the changes.

[mark-jensen]

This raises a broader issue around improving and standardizing the format of definitions for relations in CCO.

Changing axioms requires a choice to be made about deprecation. Given that we about to mint v1.6, I think this should be pushed to v1.7 or 2.0.

@avsculley Will you scan trough the rest of the temporal relations and look for similar issues, documenting them here?

[avsculley]

I'm on it.

…

________________________________ From: Mark Jensen ***@***.***> Sent: Sunday, September 15, 2024 9:54 AM To: CommonCoreOntology/CommonCoreOntologies ***@***.***> Cc: Alec Sculley ***@***.***>; Mention ***@***.***> Subject: Re: [CommonCoreOntology/CommonCoreOntologies] domain and range of 'instant before' and 'instant after' seem too permissive (Issue #480) This raises a broader issue around improving and standardizing the format of the definitions for the relations in CCO. Changing axioms requires a choice to be made about depreciation. Given that we about to mint v1.6, I think this should be pushed to v1.7 or 2.0. @avsculley<https://github.com/avsculley> Will you scan trough the rest of the temporal relations and look for similar issues, documenting them here? — Reply to this email directly, view it on GitHub<#480 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ATJNTN3TRLWVEK3KDY5SZ73ZWWGQVAVCNFSM6AAAAABOG4LJ3SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNJRGYYDGOBVHA>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[alanruttenberg]

There's a precedes/preceded-by relation in BFO that sounds like the same thing.

[swartik]

The elucidations for 'instant is before' and 'instant is after' define the properties in terms of 't1 precedes t2'. Is that plain English or a reference to the BFO property labeled 'precedes' that @alanruttenberg mentions?

[mark-jensen]

@avsculley Any progress on this?

[avsculley]

Not yet. It is next on my list. Should be done this week.

…

________________________________ From: Mark Jensen ***@***.***> Sent: Tuesday, October 15, 2024 11:19 AM To: CommonCoreOntology/CommonCoreOntologies ***@***.***> Cc: Alec Sculley ***@***.***>; Mention ***@***.***> Subject: Re: [CommonCoreOntology/CommonCoreOntologies] domain and range of 'instant before' and 'instant after' seem too permissive (Issue #480) @avsculley<https://github.com/avsculley> Any progress on this? — Reply to this email directly, view it on GitHub<#480 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ATJNTN53NVFQCVG7MF7JBMLZ3UXA7AVCNFSM6AAAAABOG4LJ3SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMJUGI3DQOJXG4>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[avsculley]

@mark-jensen I'm still working through this stuff, but here's one.

The def of 'has instide instant' is

"For Temporal Interval t1 and Temporal Instant t2, t1 has inside instant t2 if and only if there exists Temporal Instants t3 and t4 that are part of t1 and non-identical with t2, such that t3 is before t2 and t4 is after t2."

However, the domain is 'one dimensional temporal region' and the range is 'zero dimensional temporal region.'

So, either the domain and range should be changed to match the def; or the def should be changed to match the domain and range.

There is a mismatch because two disjoint instants are still a zero dimensional temporal region, and two disjoint temporal intervals are still a one dimensional temporal region.

I'm unaware of the intention behind this object property so I don't have a suggested solution.