Wikidata:Property proposal/SciProfiles ID

SciProfiles ID edit

Originally proposed at Wikidata:Property proposal/Person

Done: SciProfiles ID (P8159) (Talk and documentation)

Description	ID of an author in SciProfiles (SciProfiles (Q92645976)) by MDPI (Q6715186)
Represents	SciProfiles ID (Q92648217)
Data type	External identifier
Domain	human (Q5)
Allowed values	[1-9]\d*
Example 1	Shu-Kun Lin (Q63657451) → 2
Example 2	Franck Vazquez (Q54263688) → 66220
Example 3	Martyn Rittman (Q61863767) → 47207
Example 4	Erik Gawel (Q1353915) → 94648
Example 5	Björn Brembs (Q879571) → 92927
External links	Use in sister projects: [ar] • [de] • [en] • [es] • [fr] • [he] • [it] • [ja] • [ko] • [nl] • [pl] • [pt] • [ru] • [sv] • [vi] • [zh] • [commons] • [species] • [wd] • [en.wikt] • [fr.wikt].
Number of IDs in source	~750000 claimed profiles, potentially more than 100 million unclaimed profiles
Expected completeness	always incomplete (Q21873886)
Formatter URL	https://sciprofiles.com/profile/$1

Motivation edit

Note this is a new service launched in 11 October 2019 so there're no usage in Wikipedia yet. GZWDer (talk) 16:24, 18 November 2019 (UTC)[reply]

Discussion edit

Support David (talk) 06:12, 19 November 2019 (UTC)[reply]
Support. Nomen ad hoc (talk) 17:14, 19 November 2019 (UTC).[reply]
Question @GZWDer: Do you know why the ID's are so variable? Some are numbers, some mixed-case names, some lower case (just from your examples)? Are the ID's picked by users? Can they change? ArthurPSmith (talk) 22:54, 19 November 2019 (UTC)[reply]
Support I would also some clarification on ArthurPSmith's question @GZWDer: Iwan.Aucamp (talk) 12:11, 12 March 2020 (UTC)[reply]
- Comment @ArthurPSmith, GZWDer: I made a profile on sciprofiles.com and it is clearer now. By default you get a number assigned but you can set a custom string to be used instead. The numeric identifier still works, for Franck Vazquez (Q54263688) it is 66220 and for Martyn Rittman (Q61863767) it is 47207. I think we should disallow custom strings and just use the numeric identifiers. You can get the numeric identifier for a person from their profile links. There is actually a third identifier also in use that looks like some base64 encoded data. Iwan.Aucamp (talk) 11:47, 30 April 2020 (UTC)[reply]
- Comment We could even consider using the base64 identifiers as these work for unclaimed accounts also. Iwan.Aucamp (talk) 22:23, 30 April 2020 (UTC)[reply]
- Comment @GZWDer: I'm changing the proposal as I suggested. Iwan.Aucamp (talk) 22:23, 30 April 2020 (UTC)[reply]
- OK.--GZWDer (talk) 22:25, 30 April 2020 (UTC)[reply]
  - I updated the proposal now, only numeric IDs should be used. @GZWDer, ArthurPSmith, Nomen ad hoc: FYA Iwan.Aucamp (talk) 00:29, 1 May 2020 (UTC)[reply]
@GZWDer, ديفيد عادل وهبة خليل 2, Nomen ad hoc, ArthurPSmith, Iwan.Aucamp: Done as SciProfiles ID (P8159). Thierry Caro (talk) 02:42, 2 May 2020 (UTC)[reply]

Question about IDs edit

@GZWDer, ديفيد عادل وهبة خليل 2, Nomen ad hoc, ArthurPSmith, Iwan.Aucamp: can someone say how to find the numeric ID from a personalized ID like https://sciprofiles.com/profile/valexiev or a base64 like https://sciprofiles.com/profile/author/VktZK002QkZjS1o5cURzMHRmck1PdkFnWEczQjNQaGYxNTZoQUc5Ynd3L0NzSkg0ZU1abHdoS0J1NE5JMlpBYw== ? Preferably at https://www.wikidata.org/wiki/Property_talk:P8159 so it's saved for posterity
If unclaimed IDs are base66, shouldn't we still include them? Michel Dumontier is a well-known researcher in sem web and biomedical informatics, but I doubt he'll go claim that profile. All his links are that base64 URL, eg see https://sciprofiles.com/search/publications/0?text=michel+dumontier . BTW I tried that code at https://www.base64decode.org/ but nothing meaningful comes out. --Vladimir Alexiev (talk) 20:09, 26 May 2020 (UTC)[reply]

I wanted to leave this question at ttps://sciprofiles.com/contact buh there is some hidden (no-working) captcha that prevents this: "i! sciprofiles has great functionality, and Wikidata would like to link to you. Could you please provide some clarification about your IDs at h_here_ Also, if there's a database export that we could use to create a Mix-n-Match data catalog, that would be appreciated.

Does anyone know somebody at MDPI? --Vladimir Alexiev (talk) 20:12, 26 May 2020 (UTC) |[reply]

Do some people not have a numeric id, even if they've set up a personalized id? Is the base64 one available in all cases? ArthurPSmith (talk) 20:42, 26 May 2020 (UTC)[reply]

@ArthurPSmith: all profiles have numeric IDs also - and only numeric IDs should be used for this property. Iwan.Aucamp (talk) 07:48, 27 May 2020 (UTC)[reply]

@Vladimir Alexiev: the numeric ID for https://sciprofiles.com/profile/valexiev is 911684. You can get the numeric id from the link on the "Projects" button which is https://sciprofiles.com/projects/profile/911684 - you can also get it from the HTML which has followingButton_911684 Iwan.Aucamp (talk) 07:58, 27 May 2020 (UTC)[reply]

@Vladimir Alexiev: unclaimed profiles do not have any values that you can plug into the formatter URL (https://sciprofiles.com/profile/$1) that will resolve as far as I can tell - so if we want to use them we have to use a different formatter URL and only use the baseXX identifers for it. I'm am not sure they will be disambiguated either, so you might have multiple for one person. Iwan.Aucamp (talk) 08:17, 27 May 2020 (UTC)[reply]

Thanks, added as "syntax clarification" on the numeric regex. So @Iwan.Aucamp, ArthurPSmith, GZWDer, ديفيد عادل وهبة خليل 2, Nomen ad hoc: should we make a proposal for new prop "SciProfiles base64 hash"? It seems that most authors are still unclaimed. GZWDer, where you got the total number of records from, and is there any dump available? --Vladimir Alexiev (talk) 07:15, 28 May 2020 (UTC)[reply]

@Vladimir Alexiev: I think if there is a good use for it then we should maybe add it, but I'm a bit skeptical as it stands if the base64 identifiers are of much use over just plain text strings with ways an author's name is represented. I think if we could show that it would be useful for author disambiguation that it could be justified. Iwan.Aucamp (talk) 07:27, 28 May 2020 (UTC)[reply]

@Iwan.Aucamp, GZWDer: Given we don't have a connection to MDPI, no idea whether there is a dump, nor any idea of the stability of the base64 hashes, it's not worth the effort at this point --Vladimir Alexiev (talk) 07:30, 29 May 2020 (UTC)[reply]

Numeric vs name-based IDs edit

Notified participants of WikiProject Authority control

@GZWDer, ديفيد عادل وهبة خليل 2, Nomen ad hoc, ArthurPSmith, Iwan.Aucamp, Thierry Caro:

https://sciprofiles.com/profile/911684 now redirects to https://sciprofiles.com/profile/valexiev. So I'm reopening the question what should be the proper form of this ID.

Please comment --Vladimir Alexiev (talk) 07:00, 19 August 2021 (UTC)[reply]

I'm not sure all IDs have been redirected, e.g. https://sciprofiles.com/profile/94648; maybe we should allow both forms, although it would probably be confusing; I would wait for more investigation into the situation of the IDs. --Epìdosis 07:03, 19 August 2021 (UTC)[reply]

For discussion of already created property, please use property talk page.--GZWDer (talk) 14:12, 19 August 2021 (UTC)[reply]