Wikidata:Property proposal/present in non-fictional work

From Wikidata
Jump to navigation Jump to search

present in non-fictional work

[edit]

Originally proposed at Wikidata:Property proposal/Generic

   Not done
Descriptionlink to item which mentions this item
Data typeItem
Domainhuman (Q5) and P31/P279*:organization
Allowed valuesP31/P279*:document, P31/P279*:document series or P31/P279*:book (edition)
Example 1Madonna (Q1744)Paradise Papers (Q42722842)
Example 2Gerhard Schröder (Q2530)Paradise Papers (Q42722842)
Example 3Elizabeth II (Q9682)Paradise Papers (Q42722842)
See alsopresent in work (P1441) described by source (P1343)

Motivation

[edit]

This is useful for modeling when a company or person is mentioned in an information leak, e.g. Paradise Papers (Q42722842). At the moment significant event (P793) is widely used.--So9q (talk) 10:30, 6 October 2021 (UTC)[reply]

Discussion

[edit]

 Oppose widen the scope of present in work (P1441) instead. It already allows for non-fictional works, just not non-fictional entities, so if this was approved there would be overlap with these two properties as it is now. Ainali (talk) 10:59, 6 October 2021 (UTC)[reply]

present in work (P1441) actually recommends using described by source (P1343) for non-fictional works. That might be suitable here also though. ArthurPSmith (talk) 16:54, 6 October 2021 (UTC)[reply]

 Oppose As Ainali. present in work (P1441) allows for non-fictional works. described by source (P1343) is for cases when the work has been used as a source. /ℇsquilo 09:25, 12 October 2021 (UTC)[reply]

  • No, that would not be lost. If the work is fictional (which is something we will keep as it belongs on the item of the work) then what you think is lost is always implied through the relation and the nature of the items. Ainali (talk) 15:02, 12 October 2021 (UTC)[reply]
    • @Ainali: Wikidata does not work on the basis of implied knowledge because of the open world assumption. Changing the meaning of the property as proposed introduces an ambiguity, thus information is lost. --SilentSpike (talk) 10:06, 13 October 2021 (UTC)[reply]
      On the contrary, Wikidata have tons of data being implicit. This is what we use subclasses and all other transitive properties for in the first place. We also have many properties that could be reciprocal but are not, which means that the data is only explicit in one direction and implicit in another. This would be analogous to that. Ainali (talk) 20:38, 13 October 2021 (UTC)[reply]
      I disagree as those are both examples of explicitly captured knowledge. The subclass relation is explicitly in the knowledge base. Similarly there is no such thing as a one way statement. Every statement involving two items explicitly captures knowledge about both items even if not immediately apparent to a human via the UI (check out the relateditems gadget in preferences for a good example of this). SilentSpike (talk) 21:17, 13 October 2021 (UTC)[reply]
  •  Comment Having completely separate but otherwise identical properties for fiction and nonfiction works requires editors to decide if a work is fiction or nonfiction. For many works this would be simple and straightforward (such as the case study of the Panama/Pandora papers), but for the rest sources would be required. What would be the proposed way to convey them? Arlo Barnes (talk) 22:13, 12 October 2021 (UTC)[reply]