Wikidata:Property proposal/affiliation string

affiliation string

Originally proposed at Wikidata:Property proposal/Creative work

Done: affiliation string (P6424) (Talk and documentation)

Description	qualifier to describe the affiliation of an author in an author name string (P2093)
Data type	String
Domain	statements with author name string (P2093) property
Example 1	Registry of Open Access Repositories (Q861552) author name string (P2093) "Tim Brody" qualifier → "Southampton University"
Example 2	Ultimo - Il Capitano che Arrestò Totò Riina (Q4003222) author name string (P2093) "Maurizio Torrealta" qualifier → "italian writer"
Example 3	Space-Time Approach to Quantum Electrodynamics (Q21709213) author name string (P2093) "Richard Feynman" → "Department of Physics, Cornell University, Ithaca, New York" [fictitious: already replaced]
See also	author name string (P2093), author (P50), affiliation (P1416), educated at (P69), employer (P108)

Motivation

@Daniel Mietchen, Harej, Research Bot, James Hare, Succu: @Fnielsen, Andrawaag, Richard Nevell, Zuphilip, Egon Willighagen: @MartinPoulter, Magnus Manske, ArthurPSmith: The Source MetaData WikiProject does not exist. Please correct the name.

Potentially, this could make author name strings easier to disambiguate. When author name string (P2093) is replaced with author (P50), this can be removed. The information should be present in any of affiliation (P1416), educated at (P69), employer (P108) on the item for the author (Add your motivation for this property here.)

Ping to users who add P2093/replace them with P50 frequently. Please ping those I omitted. --- Jura 14:25, 23 December 2018 (UTC)[reply]

Discussion

I like this idea. It all starts with data. Once we have this data, we can start using a supported translation into a machine readable affiliation. --Egon Willighagen (talk) 14:42, 23 December 2018 (UTC)[reply]

Support However, can we get some examples in there? ArthurPSmith (talk) 14:02, 24 December 2018 (UTC)[reply]

Update - I added an example - using it as a qualifier on P50 instead of P2093, I assume it should be fine for either? ArthurPSmith (talk) 14:05, 24 December 2018 (UTC)[reply]

No, the proposal is just for P2093. --- Jura 08:55, 28 December 2018 (UTC)[reply]

Why? Did you check my example? How would you handle that otherwise? ArthurPSmith (talk) 19:06, 28 December 2018 (UTC)[reply]

By adding the information to Richard Feynman (Q39246) or Cornell University (Q49115) or Ithaca (Q217346). --- Jura 19:41, 28 December 2018 (UTC)[reply]

In what way could you add "Department of Physics, Cornell University, Ithaca, New York" to one of those items to indicate that was how the person's affiliation was shown on that particular paper? We can't use the object named as (P1932) qualifier on the author (P50) entry on the paper either because that is used for the way the author name is represented (and the author name string (P2093) is transformed into the object named as (P1932) qualifier when author (P50) is added). So I don't see any way around keeping this if we want to preserve these strings. ArthurPSmith (talk) 17:19, 2 January 2019 (UTC)[reply]

Presumably we don't want to preserve these strings. Using this property would be a temporary measure exclusively. The purpose is to more easily replace P2093 statements. --Yair rand (talk) 19:39, 2 January 2019 (UTC)[reply]

@Yair rand: I don't see why not - affiliation as provided in a publication often includes considerably more detail than an "employer" statement - as in the Feynman example I provided. Removing these statements or disallowing them on P50 seems to me a needless loss of useful information, and I strongly object to this needless constraint. ArthurPSmith (talk) 02:17, 21 January 2019 (UTC)[reply]

Support I added an exemple. — eru ^{[Talk] [french wiki]} 10:36, 28 December 2018 (UTC)[reply]

Support Good idea.--Jklamo (talk) 12:54, 28 December 2018 (UTC)[reply]