Wikidata

Wikidata:Property proposal/happened to

< Wikidata:Property proposal

internal change of state of edit

Note
Property renamed, was happened to

Originally proposed at Wikidata:Property proposal/Generic

   Not done
Descriptionthis event is an internal change of state of that entity. Alias : « occurs in »
Data typeItem
Domainevents that occurs to something
Example 1
⟨ French Revolution (Q6534)      ⟩  Wikidata property  ⟨ France (Q142) ⟩
Example 2
⟨ death of Diomedes Díaz (Q24943895)      ⟩  Wikidata property  ⟨ Diomedes Díaz (Q5256493) ⟩
Example 3
⟨ Big Bang theory (Q323)      ⟩  Wikidata property  ⟨ Universe (Q1) ⟩
See alsosignificant event (P793), participant (P710), facet of (P1269)

Motivation edit

(Ajoutez ici vos motivations pour la création de cette propriété) I think we need a property for items whose topic is an event that happens to change the internal state of some real world entity, to link . This is different from participant (P710)   as this property is for events like sport competition that involves several entities but the entities themselves remains the same after the event. On the other hand, a birth, a constitutional change, a disease someone happens to catch involves a single actor who is changed by the event. This property is intended only for those latter cases, and not a meeting event or other event involving several actors (sport competition, wars, …). For example the french revolution happened to the france country administration, technically, and had many participants. It changed the french political regime (the change in the state of the entity) but it did not change the actors themselves (although some actors might have a change in their own life/death status in the process). author  TomT0m / talk page 12:28, 3 October 2018 (UTC)[reply]

Notes edit

Discussion edit

  •   Comment So this would be an inverse of significant event (P793) ? Do we need both? Jheald (talk) 16:13, 3 October 2018 (UTC)[reply]
    @Jheald: Not really, it seems to me that « key event » is way more broadly used and that in many case it could be an inverse of participant (P710)   or some other property. Maybe incorrectly, but pragmatically it seems that it’s used in many other usecases (and it’s hard to catch the possible misuse, if it’s actually supposed to be used only in the circonstances of my proposal, it’s hard to catch by constraints). Just sample a few results of this query that just checks subject and objects of key event and you’ll understand what I mean. author  TomT0m / talk page 18:34, 3 October 2018 (UTC)[reply]
    To complete, this could work if, like in BFO and OBO, we had a way to know that an event type is tight to a type of entity (a death event or birth event is uniquely tight to a living entity, for example, but a war usually involves several actors). Then if we have a class « event tight to an individual », that « death » (for example) and all such event classes are a subclass of it, and that the key event value is an instance of such class, then we can compute the equivalent as the result of a query. But in Wikidata it’s a bit hard to know this from the class tree imho, and it’s still not an inverse property because « significant or notable events associated with the subject » is not a suitable description (even if this was the initial intension, don’t really remember, and if it is we went way out of the road). author  TomT0m / talk page 18:45, 3 October 2018 (UTC)[reply]
  •   Comment French Revolution (Q6534) has already a property "location"; death of Diomedes Díaz (Q24943895) P31 death of (P642): Diomedes Díaz (Q5256493). Big Bang theory (Q323) can also use "of" qualifier. Germartin1 (talk) 21:12, 3 October 2018 (UTC)[reply]
    @Germartin1: « location » definitely miss the point. The death of someone can occur on some place, this does not mean that the place is dead. That and event about france took place in france is not exactly surprising, but that it took place in france does not mean the revolution is a change of the french regime. I mentioned the « of » qualifier in my proposal, please comment on the reason I think it’s not an ideal solution. Besides, I think for a death it’s clear what « of » mean, but for other situations the qualifier might not be so clear. Imagine a heart attack of someone ? This makes barely sense in english, imagine for other kind of events. Besides it’s highly ambiguous, see its usage for a specific event as a qualifier of « instance of » in : second plague pandemic (Q18341002) (found in the first results of this query, so really not hard to find unless I’m very lucky : this is a totally different meaning ! a consistent use would imply that the pandemia occured in a human group, like a death happens to someone, here it’s used to say which kind of disease it’s a pandemia of … Terrible solution. author  TomT0m / talk page 21:38, 3 October 2018 (UTC)[reply]
  •   Support David (talk) 08:32, 4 October 2018 (UTC)[reply]
  •   Comment this duplicates participant (P710). The English description of that property seems to have a larger scope than the French one. --- Jura 09:21, 4 October 2018 (UTC)[reply]
    • @Jura1: The descriptions and name may need to be better chosen or clarified. Nevertheless the motivation for this property specifically mentions why it’s different from « participant ». A person do not « participate » in its own death like he would participate in a sport competition, it’s an event that happens to himself and only himself, profoundly change him (its « state ») (euphemism for a death). The universe do not « participate » in the « big bang », it is a change in the state of itself. (besides, it seems that the universe is not in the scope of the constraint of « participant », who is restricted to stuffs like sport competition according to current constraints). author  TomT0m / talk page 09:36, 4 October 2018 (UTC)[reply]
      • At first it may seem so, but if you qualify the statement correctly ("object has role"), it doesn't (for example #2). I'm not really sure what's the added benefit of this somewhat vague property in other cases. We already have other general properties that can be used: facet of (P1269), significant event (P793). --- Jura 11:23, 8 October 2018 (UTC)[reply]
        • @Jura1: It’s definitely not « vague », or you should say this to one of the most rigorous modelling frameworks ever creators, BFO and OBO :) On the other hand please read the description of « object has role » : « role or generic identity of subject (the item that the statement is on) in a certain context. » Seriously wtf ? What should be the role of the big bang in the universe in « a certain context » (meaning what, in the context of the statement that the big bang is a key event for the universe)? You mean an item whose title is « the object is an internal change of state of the subject ? » This is becoming to be especially abstruse. author  TomT0m / talk page 11:39, 8 October 2018 (UTC)[reply]
          • Sorry if it wasn't clear: I just inserted "(for example #2)" in my previous comment. --- Jura 11:54, 8 October 2018 (UTC)[reply]
            • @Jura1: Sorry but I see nothing in Diomedes Díaz (Q5256493). In example 1 there is even no « key event » statement in the item of france - and we can understand why, it would bloat the item. It could be that « french revolution » is part of « history of france », but I’m afraid history of France (Q7778) is an aspect of « history of europe » both with an aspect of france, good luck to sort out the meaning of both, this is a mess. And indeed, « aspect of » is a vague property, it does not explain the nature of the relation and people uses it as a vague property, unsurprisingly. Maybe you can recover your stuffs by a query « this is the history of a country, let find the country listed if the « aspect of » statement, it’s likely to be the good one. Then any part of this history is likely to have happen to that country". Good luck if you have two « aspect of » statements to countries. This is trading a clear property by way more complicated ways to do the link. (I emphasize the plural to emphasize this imply how inconsistenly and unpredictably we model similar relations) author  TomT0m / talk page
            • @Jura1: Oh, I realized you may refer to the addition of the « of » qualifier with another account (?) Then you did not answer on the fact that the « of » qualifier has many meaning and usages, even on events item which lead to ambiguities. How do you sort out all this ? author  TomT0m / talk page 13:14, 8 October 2018 (UTC)[reply]
  •   Comment_this inspires me another argument, the risk to bloat the item with « key event » for items for which happens a lot of events should motivate us to create the property. This is a functional property, each event has at most one value for it, whereas in the sense « subject -> internal change event » we have many values in one item. Such consideration made us creating no inverse property of « subclass of », because a class has less subclass than superclass (usually). Here it’s used in the other way, we should favor many statements in one item other one statement in many items ? author  TomT0m / talk page 12:22, 8 October 2018 (UTC)[reply]
  •   Oppose Use participant (P710). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:08, 8 October 2018 (UTC)[reply]
    • _@Pigsonthewing: Do you « participate » in your own death ? author  TomT0m / talk page 08:23, 11 October 2018 (UTC)[reply]
      • Not for a while I hope, but I'll be very disappointed if it happens when I'm not there. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:29, 11 October 2018 (UTC)[reply]
        • @Pigsonthewing: :) More seriously, there is a difference into a football game, a meeting event on one hand, in which several entities joins the same place to do something, but that something is external to themselves, and a heart attack, or the inflation of the universe, in which something internal happens to an entity. Using « participant » cannot reflect that difference in nature. There is probably items for which the two stuffs could be mixed up, for example imagine the modification of a forest from wild state to human engineered and exploited one. The forest itself is modified, but external entities participate in the process, humans, companies … The forest do not participate in its « artificialisation », it’s a change of its state, however. In the latter case, we can’t use « participant ». author  TomT0m / talk page 11:49, 11 October 2018 (UTC)[reply]
  •   Oppose. The various uses are already covered more specifically by other properties. --Yair rand (talk) 19:05, 9 October 2018 (UTC)[reply]
    • @Yair rand: I don’t agree ( what fits for the « Big Bang » ? ) None of these « ad hoc » « specific » solutions are exempt of defaults - see the problems exposed in my other answers) It should be seen (imho) as a problem that a very similar relationship in nature has many inconsistent and imperfect ways to be modelled on Wikidata. author  TomT0m / talk page 08:23, 11 October 2018 (UTC)[reply]
      • @TomT0m: I think it would be helpful if you could provide more examples. We already have Universe (Q1) significant event (P793) Big Bang theory (Q323), but it sounds like you're getting at the issue of this being a particular subset of types of events? If so, couldn't the target items just be placed into a "change of state" (or something like that) class tree to indicate that? (Btw, the English label of P793 is "significant event", not "key event". You might want to use {{P}} when referencing it, to avoid confusion.) --Yair rand (talk) 15:49, 17 October 2018 (UTC)[reply]
        • @Yair rand: I don’t really want to repeat myself forever, so please see my answers to similar questions from the participants above (the page is small enough for this not to be a big issue). But if only one more reason if I the others are not enough, what if an event has its own item but is not judged « significant » enough to bloat the item it happens to ? author  TomT0m / talk page 16:27, 17 October 2018 (UTC)[reply]

  Not done Not enough support.--Micru (talk) 14:09, 19 December 2018 (UTC)[reply]

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Property_proposal/happened_to&oldid=815155131"