Differentiate goesToAgent/comesFromAgent and hasRecipient/hasGiver #1152

kstudzin · 2024-08-30T14:18:25Z

Modify definitions and add examples

uscholdm

This is quite tricky. Sometimes there is no reason to have both a giver and a comes from agent, nor both a recipient and a goes to agent. In these cases, and it may not matter which to use, hence the original confusion, but we should give some guidance if possible. I would say if either seems appropriate, then use hasGiver or hasRecipient instead of the to/form agent properties. Shipment is a good example where there could be multiple parties. Let’s say a purchasing department orders a new printer and it is shipped using a third party.

Company that placed the order and who will receive the goods (recipient)
The agent the package is addressed to who will pass it along to the recipient (goes to agent)
The company that ships the package (comes from agent)
The company that is having the printer shipped (hasGiver)

In this example there might not be a separate goes to agent.

Also, I suggest that all 4 properties have a domainIncludes gist:Event

uscholdm · 2024-09-03T17:43:13Z

ontologies/gistCore.ttl

@@ -2177,6 +2177,7 @@ gist:comesFromAgent
 		) ;
 	] ;
 	skos:definition "The party that is the source of something (e.g. a message, shipment, etc.)"^^xsd:string ;
+	skos:example "For a shipment, the delivery driver is the giver and the supplier is the comes from agent"^^xsd:string ;


For the example stick with the property in question. E.g.

`For a shipment, the supplier is the comes from agent'

Maybe a more complete example with all 4 properties could be in a biggish scope note that is repeated for all 4 properties.

uscholdm · 2024-09-03T17:45:58Z

ontologies/gistCore.ttl

@@ -2612,7 +2613,8 @@ gist:goesToAgent
 			gist:Person
 		) ;
 	] ;
-	skos:definition "The party that is the recipient of something (e.g. a message, shipment, etc.)"^^xsd:string ;
+	skos:definition "The party that something (e.g. a message, shipment, etc.) is directed towards"^^xsd:string ;
+	skos:example "For a savings bond given to a minor, the minor is the goes to agent and their parent/gardian is the recipient."^^xsd:string ;


The bond example is a bit confusing to me. What is it that has a recipient? The event of giving a savings bond to a minor?

rjyounes · 2024-09-06T20:17:48Z

Also, I suggest that all 4 properties have a domainIncludes gist:Event

I don't think of Event as the likeliest domain for goesTo/comesFrom.

rjyounes · 2024-09-06T20:18:47Z

This issue has not been triaged yet, and I think as a group we need to come to some understanding of what the differences are meant to be. Moving to draft state.

uscholdm · 2024-09-06T20:45:47Z

Also, I suggest that all 4 properties have a domainIncludes gist:Event

I don't think of Event as the likeliest domain for goesTo/comesFrom.

Its the only one I could think of right away, what are the obvious ones that come to mind for you?

rjyounes · 2024-09-09T15:24:11Z

Its the only one I could think of right away, what are the obvious ones that come to mind for you?

As the definitions say, I think of shipment, package, letter, etc., which we don't have classes for. I don't see any reason that we must have a domainIncludes value, so if nothing fits the bill I'd leave it off.

uscholdm · 2024-09-09T16:42:09Z

As the definitions say, I think of shipment, package, letter, etc., which we don't have classes for. I don't see any reason that we must have a domainIncludes value, so if nothing fits the bill I'd leave it off.

@rjyounes Shipment is an event; if a package or letter goes anywhere, that is an event, which is always a correct way to model a 'going'. One need not always, model the event explicitly. But I cannot think of a single example of using goesToAgent where there is no event going on (so to speak) - can you?

rjyounes · 2024-09-09T17:57:20Z

Oh, I see, I hadn't been thinking of the events - I was thinking of the objects being shipped and delivered. The locution seems odd in ordinary English - "the event of shipping the package goes to Tom" as opposed to "the package goes to Tom." The distinctions you make above between those who handle goods in passing them from the "ultimate" giver to the "ultimate" recipient are clear, but it's not at all clear to me how we decide which set of predicates goes with which. One pair of agents are the "intermediate" givers and recipients, the other are the "ultimate" ones. I have no clue why one is "comesFrom/goesTo" and the other is "giver/recipient." Do you? Do we need to rethink our wording choices?

uscholdm · 2024-09-10T15:15:46Z

Oh, I see, I hadn't been thinking of the events - I was thinking of the objects being shipped and delivered. The locution seems odd in ordinary English - "the event of shipping the package goes to Tom" as opposed to "the package goes to Tom." The distinctions you make above between those who handle goods in passing them from the "ultimate" giver to the "ultimate" recipient are clear, but it's not at all clear to me how we decide which set of predicates goes with which. One pair of agents are the "intermediate" givers and recipients, the other are the "ultimate" ones. I have no clue why one is "comesFrom/goesTo" and the other is "giver/recipient." Do you? Do we need to rethink our wording choices?

Those are all excellent points, the locution is not good- the event does not go to the agent, the package does. But changing the wording will only work if we are rock solid on what we are trying to say - we may not be. Intermediate vs. ultimate may or may not be exactly what we want. There may be too many variations of this. I wonder if anyone ever used the comesToAgent and goesToAgent properties? I wonder if we should just remove them. Maybe a gist dev group discussion?

Add release notes

kstudzin · 2024-10-16T13:46:58Z

ontologies/gistCore.ttl

@@ -2190,7 +2194,8 @@ gist:comesFromPlace
 			gist:Place
 		) ;
 	] ;
-	skos:definition "Origin"^^xsd:string ;
+	skos:definition "Relates something in motion or something that enables motion to its place of origin."^^xsd:string ;


I included 'something that enables motion' to cover the case where the subject is gist:GeoSegment because the segment itself is not in motion.

#1165 may remove gist:GeoSegment, in which case we could consider also removing the phrase 'something that enables motion.' When we need to connect the pathway to its origin, we could use gist:linksFrom like we do with networks.

See comments on comesFromAgent.

kstudzin · 2024-10-16T13:49:06Z

ontologies/gistCore.ttl

@@ -2626,7 +2635,8 @@ gist:goesToPlace
 			gist:Place
 		) ;
 	] ;
-	skos:definition "Destination"^^xsd:string ;
+	skos:definition "Relates something in motion or something that enables motion to its destination."^^xsd:string ;


See gist:comesFromPlace comment

With #1165, we could use gist:linksTo to connect a pathway to its destination.

That's a very interesting point, and suggests that GeoRoute ought to be a subclass of Network as well as Place, which seems correct.

uscholdm

Its a bit unclear what the intended domain of these 4 properties should be.

uscholdm · 2024-10-17T14:59:32Z

ontologies/gistCore.ttl

@@ -2176,9 +2176,13 @@ gist:comesFromAgent
 			gist:Person
 		) ;
 	] ;
-	skos:definition "The party that is the source of something (e.g. a message, shipment, etc.)"^^xsd:string ;
+	skos:definition "Relates something in motion to the agent that originated the motion."^^xsd:string ;


This property is intended for use with (among otherr things) a shipping event, which involves movement of a package, but the event itself is not moving. It would be good to include intended domain here. It is tempting to say a physical thing, such as the TV I just ordered from Amazon, but the agent and place where the tv cam and went from/to is not a property of the TV, but of the shipping event.

This definition places the focus on the motion rather than the origin. It also would be "is or has been" in motion. I would limit the definition to the concept of source/origin, and add the part of movement through space to a scope note that explicitly differentiates it from hasGiver. I also think the phrase "physical or electronic transfer" might work better than "in motion."

uscholdm · 2024-10-17T15:04:56Z

ontologies/gistCore.ttl

 	skos:prefLabel "goes to agent"^^xsd:string ;
-	skos:scopeNote 'This is not the inverse of gist:comesFromAgent. A message can be from someone. If we made it the inverse the person would be "to" the message'^^xsd:string ;
+	skos:scopeNote
+		"In general, 'gist:goesToAgent' is intended for when the subject moves. However, there is some overlap in when 'gist:hasRecipient' and 'gist:goesToAgent' can be used, such as recieving a gift."^^xsd:string ,


I think this it a good way to handle the overlap when either seems appropriate.

rjyounes

So far I've commented in detail only on one of the properties, comesFromAgent, and I recommend similar updates for the others. I can re-review once those have been made.

docs/release_notes/issue1024-differentiation-giver-sender-props.md

rjyounes · 2024-10-17T15:55:19Z

ontologies/gistCore.ttl

@@ -2176,9 +2176,13 @@ gist:comesFromAgent
 			gist:Person
 		) ;
 	] ;
-	skos:definition "The party that is the source of something (e.g. a message, shipment, etc.)"^^xsd:string ;
+	skos:definition "Relates something in motion to the agent that originated the motion."^^xsd:string ;


This definition places the focus on the motion rather than the origin. It also would be "is or has been" in motion. I would limit the definition to the concept of source/origin, and add the part of movement through space to a scope note that explicitly differentiates it from hasGiver. I also think the phrase "physical or electronic transfer" might work better than "in motion."

ontologies/gistCore.ttl

rjyounes · 2024-10-17T16:03:59Z

ontologies/gistCore.ttl

@@ -2190,7 +2194,8 @@ gist:comesFromPlace
 			gist:Place
 		) ;
 	] ;
-	skos:definition "Origin"^^xsd:string ;
+	skos:definition "Relates something in motion or something that enables motion to its place of origin."^^xsd:string ;


See comments on comesFromAgent.

…s.md Co-authored-by: Rebecca Younes <[email protected]>

Co-authored-by: Rebecca Younes <[email protected]>

uscholdm

We are nearly there - tough nut to crack. I 'm thinking maybe replace

In general, this property is used in abstract contexts with
This property is often used in abstract contexts

The former may be too strong, but that would mean rephrasing uses of 'typically' such as

gist:hasGiver is typically used in more abstract contexts such as agreements, obligations, contract
We may be better off leaving it as it is.

kstudzin · 2024-10-23T18:22:38Z

@uscholdm I made that change - let me know what you think

kstudzin requested review from rjyounes and uscholdm August 30, 2024 14:18

kstudzin self-assigned this Aug 30, 2024

uscholdm requested changes Sep 3, 2024

View reviewed changes

rjyounes marked this pull request as draft September 6, 2024 20:18

Update defs, examples, and scope notes

b5471ae

Add release notes

kstudzin force-pushed the kate/1024 branch from 574b094 to b5471ae Compare October 16, 2024 13:36

kstudzin commented Oct 16, 2024

View reviewed changes

kstudzin requested a review from uscholdm October 16, 2024 13:49

uscholdm requested changes Oct 17, 2024

View reviewed changes

rjyounes requested changes Oct 17, 2024

View reviewed changes

kstudzin and others added 4 commits October 17, 2024 13:40

Update docs/release_notes/issue1024-differentiation-giver-sender-prop…

0342f9c

…s.md Co-authored-by: Rebecca Younes <[email protected]>

Apply suggestions from code review

049d7cf

Co-authored-by: Rebecca Younes <[email protected]>

Updates based on review comments

e1c2407

Fix typo

08efc0a

kstudzin requested review from rjyounes and uscholdm October 17, 2024 19:19

uscholdm approved these changes Oct 18, 2024

View reviewed changes

Scope note rewording

046498b

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Differentiate goesToAgent/comesFromAgent and hasRecipient/hasGiver #1152

Differentiate goesToAgent/comesFromAgent and hasRecipient/hasGiver #1152

kstudzin commented Aug 30, 2024

uscholdm left a comment •

edited by rjyounes

Loading

uscholdm Sep 3, 2024

uscholdm Sep 3, 2024

rjyounes commented Sep 6, 2024

rjyounes commented Sep 6, 2024 •

edited

Loading

uscholdm commented Sep 6, 2024

rjyounes commented Sep 9, 2024

uscholdm commented Sep 9, 2024 •

edited

Loading

rjyounes commented Sep 9, 2024 •

edited

Loading

uscholdm commented Sep 10, 2024

kstudzin Oct 16, 2024 •

edited

Loading

rjyounes Oct 17, 2024

kstudzin Oct 16, 2024

rjyounes Oct 17, 2024

uscholdm left a comment

uscholdm Oct 17, 2024

rjyounes Oct 17, 2024

uscholdm Oct 17, 2024

rjyounes left a comment

rjyounes Oct 17, 2024

rjyounes Oct 17, 2024

uscholdm left a comment

kstudzin commented Oct 23, 2024

Differentiate goesToAgent/comesFromAgent and hasRecipient/hasGiver #1152

Are you sure you want to change the base?

Differentiate goesToAgent/comesFromAgent and hasRecipient/hasGiver #1152

Conversation

kstudzin commented Aug 30, 2024

uscholdm left a comment • edited by rjyounes Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rjyounes commented Sep 6, 2024

rjyounes commented Sep 6, 2024 • edited Loading

uscholdm commented Sep 6, 2024

rjyounes commented Sep 9, 2024

uscholdm commented Sep 9, 2024 • edited Loading

rjyounes commented Sep 9, 2024 • edited Loading

uscholdm commented Sep 10, 2024

kstudzin Oct 16, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

uscholdm left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rjyounes left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

uscholdm left a comment

Choose a reason for hiding this comment

kstudzin commented Oct 23, 2024

uscholdm left a comment •

edited by rjyounes

Loading

rjyounes commented Sep 6, 2024 •

edited

Loading

uscholdm commented Sep 9, 2024 •

edited

Loading

rjyounes commented Sep 9, 2024 •

edited

Loading

kstudzin Oct 16, 2024 •

edited

Loading