Wikidata talk:WikiProject Parenthood

From Wikidata
Jump to navigation Jump to search

relatives[edit]

Copy from project chat: "What about using human relationship properties in entities describing the relationship for example setting father (P22) in father (Q7565) to paternal grandfather (Q19682162)". Paweł Ziemian (talk) 20:28, 14 March 2016 (UTC)[reply]

While I think it's a good idea to provide various mappings of relationship aspects, I was a bit doubtful about the approach. It seems to work out, even on https://tools.wmflabs.org/reasonator/geneawiki2/?q=Q7565
I suppose we need an item to start out with: Q23065529. Not sure how to call it.
--- Jura 22:05, 15 March 2016 (UTC)[reply]
Should "grandfather" be:
  1. the "father" of "parent" only
  2. also the "father" of "mother" and "father"?
Similar to (2), there are items that could go in dozens of places.
--- Jura 22:34, 15 March 2016 (UTC)[reply]
  • There is a question if it is possible to use relative (P1038) with kinship to subject (P1039) to describe more complex relations. I you look at the simple properties, which I added like in the example above, the property is set from child perspective. So assuming that child sees the father then father's children can be child's brother or sister. Going this way the pattern for relative (P1038) is: father's relative (P1038) (child's kin) kinship to subject (P1039) (father's kin). The example can be:
⟨ father (Q7565)  View with Reasonator View with SQID ⟩ relative (P1038) View with SQID ⟨ stepbrother (Q20746702)  View with Reasonator View with SQID ⟩
kinship to subject (P1039) View with SQID ⟨ stepson (Q4346792)  View with Reasonator View with SQID ⟩

However this became very difficult to understand. Going back to simple properties, if the item is i.e. daughter (Q308194), then the view point is father (Q7565) or mother (Q7560), so the spouse (P26) can be assigned with son-in-law (Q10942739). But for complex relations it is not obvious what is the view point. Paweł Ziemian (talk) 22:53, 15 March 2016 (UTC)[reply]

For items on Q23065529, it seems to give a good overview, but, if we keep this, we will have to redo all constraints for properties.
--- Jura 00:15, 16 March 2016 (UTC)[reply]
  • Sleeping over this, finally, I wonder if we shouldn't attempt to create a separate property, just like "relative" to define these mappings. This way the basic properties can be left to actual people (and animals and .. ).
    --- Jura 21:50, 16 March 2016 (UTC)[reply]

Please look at those two examples Kapiolani (Q3322857) and Benoît Fould (Q2896518) which use two different ways to use the qualifier kinship to subject (P1039) with value adopted child (Q25858158). The former is using this qualifier on relative (P1038), and the latter on child (P40). Those two options are used by a similar number of cases (less than 50 in both cases). In the linked discussion, I first defended the former option, but thinking twice, I feel better to have adopted child represented as a child with qualifier(s) rather than a relative.
I would like to have consensus here on either option in order to edit those cases to have a consistent representation. Bear in mind that representing adopted child/adoptive parent as a qualified child/parent relationship entails relaxing the unicity constraint on father (P22) and/or mother (P25) in the case where both biological and adoptive parent is known, like in the children of Benoît Fould (Q2896518), see e.g., Flore Singer (Q3073858) who has currently two fathers. Bvatant (talk) 13:20, 18 August 2018 (UTC)[reply]

  • There may not be a general answer: scope and forms of adaption vary across time and countries. In general, I'd continue using P22, P25. BTW The constraint isn't mandatory, i.e. there can be exceptions.
    --- Jura 12:15, 26 August 2018 (UTC)[reply]

opposite of[edit]

Property opposite of (P461) is ambiguous. Compare grandson (Q11921506) where grandfather (Q9238344) and grandmother (Q9235758) is stored vs granddaughter (Q19756330) where grandson (Q11921506) is stored. This is a contradiction with property definition, which is expected to be symmetric. In the talk page of the property there is proposal to use criterion used (P1013) as qualifier, which lacks here very much. List of acceptable criteria might be useful, where the only obvious item is dedicated gender binary (Q5530970) recommended in the example of the qualifier. Paweł Ziemian (talk) 21:10, 17 March 2016 (UTC)[reply]

Qualifier criterion used (P1013) can help to indicate why they are considered "opposites".
--- Jura 08:24, 18 March 2016 (UTC)[reply]

Q23065529[edit]

outline of self (Q23065529) is hitting a lot of constraints. Can this be fixed? Sjoerd de Bruin (talk) 20:13, 18 March 2016 (UTC)[reply]

This is strange item itself. What is its necessity? --Infovarius (talk) 09:02, 20 March 2016 (UTC)[reply]
This item in only used on parent (Q7566) which also hits a lot of constraints. mother (P25), child (P40) etc should imho only be used to link to single persons and not to classes of persons. --Pasleim (talk) 16:16, 23 March 2016 (UTC)[reply]
See the discussion above: #relatives
--- Jura 06:17, 24 March 2016 (UTC)[reply]

WD:PC#How to connect parent (Q7566) and offspring (Q239526) etc. and others?[edit]

Just for reference: I've opened a discussion on the project chat, perhaps continuing #opposite of above. --Marsupium (talk) 09:16, 5 September 2017 (UTC)[reply]


Mapping relationships[edit]

Please see Wikidata:Property_proposal/kinship_equivalent for new ways to map relationships. Thanks.
--- Jura 13:53, 1 October 2017 (UTC)[reply]

Numbering generations and degrees of kinship[edit]

Maybe we could try to find a way to add this to the items in more detail. Currently there are some attempts, e.g. second-degree relative (Q24882651) (Special:WhatLinksHere/Q24882651). It's used with subclass of (P279) on uncle or aunt (Q21073936). We could probably improve on that as not all of them are 2nd degree relatives. Maybe P31 should be used instead or we could try to use some numeric property in combination with a qualifier to indicate which way was used to determine it.
--- Jura 18:44, 12 October 2017 (UTC)[reply]

New report: missing parent[edit]

It's at Wikidata:WikiProject Parenthood/reports/missing parent.
--- Jura 14:24, 15 October 2017 (UTC)[reply]

New reports: combine[edit]

I started a report that lists people where spouse (P26) = sibling (P3373). It should show mostly errors and Ancient Egyptians.
--- Jura 20:50, 18 October 2017 (UTC)[reply]

New reports: completeness[edit]

It's at Wikidata:WikiProject Parenthood/reports/completeness. Still needs some fine tuning.
--- Jura 12:10, 23 October 2017 (UTC)[reply]

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata_talk:WikiProject_Parenthood&oldid=734232978"