Wikipedia talk:WikiProject Unique Identifiers

Authority control

Wikipedia:Authority control was championed by the German Wikipedia at the request of their library community. It seems to have achieved goals similar to some of the goals of this project. If other project members agree, I will contact them and ask for their input. Thanks, — John Harvey, Wizened Web Wizard Wannabe, ^{Talk to me!} 19:57, 14 February 2012 (UTC)[reply]

Thanks. By all means start talking to them; it would be good to know that's been happening, and, especially whether they've engineered means by which 3rd parties can access wikipedia via UIDs. Meanwhile I'm vaguely aware of contacts being made with the UK library community by the UK chapter, and have a reasonable expectation that it'll start to feed in here once better established. I think AndyM is familiar with the Authority Control area - I saw him add somesuch refs to an article earlier today - [1]. So it's clear that authority control is well on the radar; and that this project is in need of the widest input we can get. --Tagishsimon (talk) 20:10, 14 February 2012 (UTC)[reply]

I've invited Kam Solusar who is involved with both the German and the English authority control projects. I'm hopeful, — John Harvey, Wizened Web Wizard Wannabe, ^{Talk to me!} 23:19, 14 February 2012 (UTC)[reply]

Hi, sorry that it took a while to respond, my job kept me pretty busy during the last week.

The German Wikipedia started including the PND number, the identifier of the German National Library's Personennamendatei, a few years ago. We later also added other identifiers such as LCCN, VIAF and some other identifiers from the German National Library.

As far as making articles accessible via such authority record IDs, de:User:APPER for example created a redirect tool on the toolserver that resolves the PND to the corresponding biographical article. [2] for example redirects to Albert Einstein. APPER also provides a text file ([3], updated daily) that contains all articles tagged with a PND. The German National Library uses this to add links to WP articles to their PND database (see [4] for example, "Zugehöriger Artikel in Wikipedia" (="corresponding article on Wikipedia") on the right side of the page). The German Federal Archives also use the PND numbers to add links to biographical WP articles to their website. AFAIR, the inclusion of PND numbers in Wikipedia articles was one of the reasons why the Federal Archives agreed to upload a large amount of historical photos under a free license to Commons. Some other German library networks also added links to Wikipedia articles of authors to their online databases (though I don't know which source they used to map their database entries to our articles).

One IMHO very interesting thing are BEACON files, see meta:BEACON. Such files can be used to map Wikipedia articles to corresponding pages on other websites/databases using identifiers such as the PND number. Imagine if something like this could be used (once the authority control template sees more widespread use on the English WP) by the OCLC (WorldCat, VIAF) or even the Library of Congress to easily add links to WP articles to their database entries. Regards, --Kam Solusar (talk) 05:22, 20 February 2012 (UTC)[reply]

UID access to wikipedia articles

The big idea is, we make wikipedia articles available via a UID in a systematic fashion. We already carry all sorts of UIDs on all sort of articles, generally within an infobox or via a template such as {{Authority control}}. I think we want, as a minimum and as a first implementation, to be able to create redirect pages to articles named something along the lines of /wiki/uid/[UID_Type]/[Instance of UID], such as /wiki/uid/uniprot/12345678. Presumably this could be done by a bot working through the "what links here" list for templates. We need to discuss:

Aim: support inbound links to wikipedia via UID
Proposed implementation: create redirects with UID names
Method: Bot
Naming convention: /wiki/uid/[UID_Type]/[Instance of UID]

We also need to discuss whether we need to do an RfC before we start doing this, or whether the VPP thread which kicked this project off - Wikipedia:Village_pump_(proposals)/Archive_86#UID_interface_to_Wikipedia - is sufficient. Thanks. --Tagishsimon (talk) 21:02, 14 February 2012 (UTC)[reply]

This is fairly trivial (and quite a good thing to do). But... should it not be DKFZp686J1831 or ISBN 0-596-00027-8 rather than uid/HUGO/DKFZp686J1831 and uid/ISBN/0-596-00027-8 ? Rich Farmbrough, 14:32, 14 March 2012 (UTC).[reply]

I'm open to any convention that works, but I think the namespace we'd need for, for instance, railway station idents (being TLAs) may well be taken, which points to the need to have a prefix of some sort. Including the [UID Type] in the prefix guards against non unique UIDs across a couple of different classes of UIDs. --Tagishsimon (talk) 14:38, 14 March 2012 (UTC)[reply]

Why on WP?

It seems we're talking of two separate ideas. The inbound links would seem to be a sort of Uniform resource identifier, targeted at each specific Wikipedia article. Standardizing this should be a fairly simple proposition to address, though we'd have to consider what behaviour is desired under pagemoves, deletions, redirections, and even simple version changes. Do we want really want to encourage stable inbound links to specific versions?

Outbound links, on the other hand, are very diverse, and not necessarily targetted at a specific web page. An ISBN, for instance is mapped via our booksources page to many different databases which then map that ISBN to (sets inconsistent of) metadata about one (or sometimes more) editions of a book. I'd be reluctant to see much wp effort go into this. Other projects such as Open Library are positioned to do it better. LeadSongDog come howl! 06:33, 15 February 2012 (UTC)[reply]

Right, so ISBN is already well catered for. What about other UIDs, that are not, yet? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:28, 17 February 2012 (UTC)[reply]

ISBN would be particularly impractical for our purposes because a) circa 95% of ISBNs won't resolve to Wikipedia articles at all, meaning a lot of inefficient work; b) most Wikipedia articles including ISBNs do so as citations not as identifiers of the topic, which muddies the waters for tracking; and c) there is no single usable database of ISBN > author/title lookups we can build a system from.

Not sure what would be the best one to begin working with, though - what's easily available and already widely-covered in Wikipedia? Shimgray | talk | 21:04, 19 February 2012 (UTC)[reply]

Regarding ISBNs, I was responding to your comment about outbound links. Though its a side-issue, we do need to find a way of indicating that a particular ISBN (or other UID) identifies the subject of the article. In most (but not quite all) infoboxes, for example, that is the case. As for low-hanging fruit, TagishSimon and I have already been discussing NaTPAN codes for Uk railway stations. OpenStreetMap already records them, for example, and I know of a couple of on-line services which use them as identifiers, and which could potentially include Wikipedia links based on them. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:12, 20 February 2012 (UTC)[reply]

NaPTAN codes seem like an excellent example of something we can do well here.

Regarding ISBNs, the problem is that infoboxes tend to include one or two at most - however, the more significant a work is (and thus the more likely we are to have an article on it), the more likely it is to have been widely published and have a correspondingly large number of ISBNs - querying LibraryThing for Great Expectations returns 360 known editions with assigned ISBNs, Pride and Prejudice just shy of 600. It's very unlikely we'll ever have reasonably complete lists of these in the articles, even were it desirable. An ISBN-to-article resolver would be great, but I'm not sure how we could make it work well - the data is messy and encumbered. Shimgray | talk | 21:44, 20 February 2012 (UTC)[reply]

Yes this is true. There's also a presentation layer with ISBN's which we are discussing part of at Wikipedia talk:ISBN, both in terms of targets entered to WP and search engine smarts. I've noticed that Google seems to de-prioritise WP wrt ISBNs, which is probably sensible, if not correct. (It's maybe not correct because WP is more likely to lead you to free versions of the book or alternative free online resources than a bibliographic or book chandler site.) There is a possibility of having some more detailed resource page like ISBN 2493842355. Rich Farmbrough, 13:24, 15 March 2012 (UTC).[reply]

Wrong target

I think that this WikiProject has the wrong target. In brief:

Articles already have URLs. I'm deeply skeptical that we can do better.
I don't believe that article topics have a single identity to identify. At the very least a logic would be required to disambiguate Human, List_of_Catholics, Category:Catholics, Kennedy family and John F. Kennedy and track that identity through moves, renames, history merges, deletions, interwiki translations, commons galleries, etc.
There are articles which are created as sub-stubs then hijacked to be about someone / somewhere else of the same name. There is currently no policy against this.
a UID that only works with a handful of science projects isn't going win the wide support that would be required to get even a relatively small policy change.

What I do believe is:

That it is worth tracking in human and machine comprehensible manner when a wikipedia topic is the same topic as something else.
- An new argument to Template:Cite with the semantics this citation URI points to a thing with the same topic as this article without actually saying what the subject is. The collection of URIs then become identifiers to identify the topic of the page.

Stuartyeates (talk) 08:25, 15 February 2012 (UTC)[reply]

Having read though the above, I'm not clear what your objections are.

Of course articles have URLs. how is an external system that has an ISBN number, NaPTAN code or whatever, to know at what URL the relevant article resides? Conversely, how is not including such a code in the article at a given URL helpful to people looking for (or third=party systems wanting to retrieve) further details of the subject, on an such an eternal system?
Many articles do have a single topic. Topics covered in multi-topic articles still have UIDs, which we can use for outbound links and to tell parsers what topics are discussed on a given page.
So?
Which UID "only works with a handful of science projects"? What policy must change, before we may use it?

I'm unable to parse your unnumbered bullet point; but this project is about using eternally-generated UIDs, not creating our own. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:39, 17 February 2012 (UTC)[reply]

Rephrase

Clearly there is a misunderstanding above. I'll attempt to restart my point(s). The proposal is not sufficiently clearly stated, in particular:

When, how and by whom would the link between a page and an identifier get created?
When, how and by whom would links get maintained as the topic of an article changes?
When, how and by whom would links get maintained as the topic of the identifier changes?
What specific technical infrastructure would be needed involved at the wikipedia end?
Are there reasons why current support for identifiers (and we have quite a bit of support for ISBNs, DOIs, URLs, etc) can't be extended to do this job?
Are there reasosns why this has to be done as part of wikipedia, rather than as a layer of abstraction on top of wikipedia?

Stuartyeates (talk) 06:01, 20 February 2012 (UTC)[reply]

1-4 are all questions this project intends to address, rather than for which we claim to have the answers.

5 - what current support do we have, for external services which know a UID to use it to find the relevant Wikipedia article?

6 - Who is going to create such a layer? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:05, 20 February 2012 (UTC)[reply]

5 We have a large number of templates which take identifiers (Template:Doi, Template:Lacking ISBN, Template:MeSH, Template:LCCN and even the humble Template:Cite) which are widely used and deeply entrenched in Wikipedia practise. If seems to me that some of these could have an extra option added to them to indicate that this reference to an identifier is a defining reference (i.e. the identifier points to a thing with the same topic as the wikipedia page) then per-template extractors, modelled on the PND one (see [5]) extract that into the semantic web. If you're after straight redirects, it should be simple enough to build a redirecter based on that.

6 There are are already a number of layers DBpedia being the most well known. Stuartyeates (talk) 19:29, 20 February 2012 (UTC)[reply]

1 - To give a couple of examples, I'd anticipate that UIDs found in infoboxes would be reused, perhaps by a bot, to provide a redirect

2 - I don't anticipate that articles on things like a railway station, an element, a protein, would change from being about a railway station to something else.

3 - The thing here is to use UIDs which are stable. As many of them are.

4 - I'm thinking a bot, perhaps linked to a database.

5 - No. On the contrary, we should reuse existing identifiers exactly as you go on to suggest.

6 - No. But turn the question around: is there any advantage in making redirects - which is what I foresee - part of something else than wikipedia? --Tagishsimon (talk) 20:29, 20 February 2012 (UTC)[reply]

6 DBpedia changes are not subject to the vulgarities of wikiepdia consensuses, in particular around what kinds of edits may be made by a bot; pressure to keep wikisyntax as simple as possible to avoid discouraging newbies; and focus on human editing of textual articles at the expense of machine editing of database-like articles. Stuartyeates (talk) 21:45, 20 February 2012 (UTC)[reply]

I'm always somewhat suspicious when a suggestion to do something in wikipedia is met with a response along the lines of "why not do it on this website, instead". I confess that though I've been around these parts of years, I have not come across DBpedia. For now, I think I'll slum it in the vulgarities of wikiepdia; though I'd be very pleased to see DBpedia implement much the same system as I'm envisaging for wikipedia, since it would seem to be exactly what it's there for. --Tagishsimon (talk) 22:22, 20 February 2012 (UTC)[reply]

"why not do it on this website, instead" is often a sign that WP:NOT is close and someone is trying to be constructive. YMMV. Stuartyeates (talk) 00:30, 21 February 2012 (UTC)[reply]

Well, yes, my M does V. Is WP:NOT close in this case? Are you objecting to the idea of Wikipedia having UID redirects to articles? If so, why? --Tagishsimon (talk) 13:52, 21 February 2012 (UTC)[reply]

I think being able find wikipedia articles by redirects is a great idea. I am, however, deeply sceptical that the current proposal will achieve the necessary consensus to achieve the substantial changes in wikipedia practice and/or infrastructure needed to have those redirects inside wikipedia. Stuartyeates (talk) 19:15, 21 February 2012 (UTC)[reply]

A project of interest to you

I'm a member of WP:NRHP, which relies a bit on a set of unique identifiers for the places in our project scope; thought you'd want to be aware of it. We've been increasing our dependence on them thanks to Multichill, who has helped to push through changes to make our lists more bot-accessible. Nyttend (talk) 15:29, 9 March 2012 (UTC)[reply]

Thanks Nyttend. How are you making use of UIDs in the project? --Tagishsimon (talk) 18:16, 9 March 2012 (UTC)[reply]

The National Park Service maintains a database with minimal information on each location that is listed on the National Register of Historic Places; these locations are our scope. Among the data included in the database (and just about the only thing that's included for every site, besides the name) is a reference number, which is unique. We include them in our infoboxes (see the "NRHP Reference #" line in the infobox of Ethel S. Roy House, for example), and our lists (example) have them as hidden parameters in the templates that we use to create the tables on the lists; while the number does not appear when you see the page itself, they're in the code to enable bots to harvest the information more easily. Nyttend (talk) 05:08, 10 March 2012 (UTC)[reply]

IETF language codes (and more)

These are coded in the {{ISO 639 name}} family of templates (I didn't know about the IETF standard at that point!). There's a similar series, indeed a similar set of series for {{ISO 3166 name}} and codes. Rich Farmbrough, 03:53, 24 March 2012 (UTC).[reply]

(Just to be clear we only have templates for (most of) the codes we use or have used, not for the "complete" set. The series also supports variations in a typical forgiving manner.) Rich Farmbrough, 03:54, 24 March 2012 (UTC).[reply]

BNF links

Is there any reason not to add the option of links to BNF "Notice d'autorité personne" listings in Template:Authority control? Here's an example:

http://catalogue.bnf.fr/ark:/12148/cb148025499/PUBLIC

The Commons version of this template supports it. We have plenty of biography articles about French people, where it would be particularly relevant to add this link. --Robert.Allen (talk) 00:41, 27 March 2012 (UTC)[reply]

Probably not, but that's not really in the scope of this project; you'd best ask on that template's talk page. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:10, 28 March 2012 (UTC)[reply]

Giving a VIAF number allows redirects via their website to other national libraries worldwide, so there's not really a reason to spread out in that direction. The additional numbers could actually be automatically pulled from the VIAF API, but there's no real reason to here. en.wiki should use truly 'universal' and English language UIDs. fr.wiki people could merge the VIAFs across (since they are truly universal) and add the french language ids. Revent (talk) 07:29, 23 April 2013 (UTC)[reply]

Time to move forward

It's gone quiet here. It would be good to have something to report at Wikimania in July.

Shall we move forawrd? Maybe with a pilot for one UID scheme? Which?

Should we - can we - link to "curid" URLs like http://en.wikipedia.org/wiki/?curid=52780 (that's an alias for U2). We might, therefore, redirect, say, UID/musicbrainz/artist/a3cb23fc-acd3-4ce0-8f36-1e5aa6a18432 (from [6]) to that.

What else is holding us up? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:23, 20 May 2012 (UTC)[reply]

a) I think curid is more or less what we want (as "number identifying specific content regardless of page moves or edits"); it may be broken if the page is deleted/recreated, but this is perhaps unavoidable.

b) Since we started discussing this, Wikidata has become an active and ongoing thing; phase 1 of that is building a single multilingual "entity database" to tie together (for example) a single entry for 76 language editions of U2, and it may well be that UIDs can (should?) be implemented at that level rather than on a single-wiki basis.

c) I'm working on a proposal involving the new! exciting! not-quite-live-yet! ISNI identifiers, but it's still a draft on my desktop and I haven't posted it yet. So there is something going on in the background ;-) Andrew Gray (talk) 12:01, 24 May 2012 (UTC)[reply]

a) Mediawiki does so well to sidestep the classic annoyance of the web by having created a system that never has broken links. It'd be a shame to loose that now. And it might not be necessary. The answer to me lies in b)

b) Wiki data will provide a new entity database that will link to other language wikis. Each page on language wiki X should have a unique identifier at least at the SQL level. The wikidata database for each entity will contain a list of the manifestations of that identity on different languages. So item "n" in wikidata should be a list of <x,y,z> where x,y, and z are the UIDs in languages X,Y, and Z.

c) here's where the magic could come in. Since all the people that will have ISNI assignments are a subset of all entities online, we can and should make a entry in entity "k" in wikidata that is an optional ISNI field. Such that k would be a list with <x,y,z,isni-id>, and this would link isni-id to each x,y and z. Maximiliankleinoclc (talk) 15:35, 30 May 2012 (UTC)[reply]

Authority control RfC

Hi all.

There's a currently open RFC on a proposal to extend the use of authority control identifiers on enwiki. It was briefly mentioned above, and seems broadly relevant to this project. Please see the proposal here and comment here. Andrew Gray (talk) 10:29, 3 July 2012 (UTC)[reply]

OIDs and URNs

Hello there. It's great to discover this WikiProject. Are you guys familiar with object identifiers (OIDs)? I'm not at all sure how prevalent they are in the world, but it seems like they would fall naturally under the scope of this project.

Additionally, I feel like we should connect to URNs in some fashion. I don't currently know how, but it's something I'll give some thought to. And maybe people here could suggest some ideas. — Hex (❝?!❞) 15:36, 7 December 2012 (UTC)[reply]

UIDS for Gene Wiki articles

A bot request to create systematic redirects for Gene Wiki articles so that they are easier to located has been submitted here. Comments and suggestions are welcome. Boghog (talk) 20:59, 1 January 2013 (UTC)[reply]

ISBN Searching

I'm trying to get some discussion going concerning some apparent limitations in the way ISBNs are currently stored on WP - not sure where to go with this, but I noticed this project and thought there might be someone interested - apologies if this is the wrong place or it is just a rehash of old material. - see Wikipedia_talk:ISBN#ISBN_Searching ---- _nonsense ferret 22:03, 2 February 2013 (UTC)[reply]

ISMN

Is there a need to create an ISMN template for music publications? Unlike ISBN, it is very easy to go to the Agency's web site and find the actual publisher based on a number. The format itself is very similar to an ISBN. --Very trivial (talk) 23:03, 3 September 2013 (UTC)[reply]

OpenCorporates and DUNS

I just came across OpenCorporates and its associated blog, where I found this interesting post on DUNS numbers. Does this tweak any interest here? I note too that the OpenCorporate data is published under the SA-BY-Open Database License (details here), we'd want to ascertain whether that's compatible with our uses. LeadSongDog come howl! 13:13, 16 October 2013 (UTC)[reply]

Mysterious LCCN error

I guess this is the right place to ask this. I just added {{Authority control}} to Peter Howitt (economist). The LCCN, however, returned an error: The LCCN id n/86/29896‏ is not valid. This is the corresponding LCCN file. What is wrong here? --bender235 (talk) 19:23, 2 December 2013 (UTC)[reply]

Fixed. Correct form was |LCCN=n86029896 LeadSongDog come howl! 22:32, 3 December 2013 (UTC)[reply]

Authority control in biographies

Should we err on the side of inclusion or exclusion of a VIAF number when the VIAF database is nebulous. We may not have birth and death years in our biographies for more obscure people, or the VIAF database does not have the birth and death years, especially when the name is not unique. --Richard Arthur Norton (1958- ) (talk) 01:14, 8 January 2014 (UTC)[reply]

Comment on the WikiProject X proposal

Hello there! As you may already know, most WikiProjects here on Wikipedia struggle to stay active after they've been founded. I believe there is a lot of potential for WikiProjects to facilitate collaboration across subject areas, so I have submitted a grant proposal with the Wikimedia Foundation for the "WikiProject X" project. WikiProject X will study what makes WikiProjects succeed in retaining editors and then design a prototype WikiProject system that will recruit contributors to WikiProjects and help them run effectively. Please review the proposal here and leave feedback. If you have any questions, you can ask on the proposal page or leave a message on my talk page. Thank you for your time! (Also, sorry about the posting mistake earlier. If someone already moved my message to the talk page, feel free to remove this posting.) Harej (talk) 22:48, 1 October 2014 (UTC)[reply]

International Standard Name Identifier (ISNI)

How does the International Standard Name Identifier fit into this approach?

ORCID is a subset of it. Non-ORCID ISNIs aren't going to be helpful on WP unless/until there's a database to link to to look them up, as there is with WorldCat as used by {{ISBN}} and {{ISSN}} (and the associated citation template parameters). It would be pretty awesome, though. — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼ 22:20, 23 October 2015 (UTC)[reply]

WikiProject X is live!

Hello everyone!

You may have received a message from me earlier asking you to comment on my WikiProject X proposal. The good news is that WikiProject X is now live! In our first phase, we are focusing on research. At this time, we are looking for people to share their experiences with WikiProjects: good, bad, or neutral. We are also looking for WikiProjects that may be interested in trying out new tools and layouts that will make participating easier and projects easier to maintain. If you or your WikiProject are interested, check us out! Note that this is an opt-in program; no WikiProject will be required to change anything against its wishes. Please let me know if you have any questions. Thank you!

Note: To receive additional notifications about WikiProject X on this talk page, please add this page to Wikipedia:WikiProject X/Newsletter. Otherwise, this will be the last notification sent about WikiProject X.

Harej (talk) 16:57, 14 January 2015 (UTC)[reply]

WP:ISSN

I've rewritten the weird anti-ISSN page Wikipedia:ISSN to be a proper {{Wikipedia how-to}} on ISSN use, and when ISSNs are especially helpful. — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼ 22:17, 23 October 2015 (UTC)[reply]

RFCs on citations templates and the flagging free-to-read sources

See

Headbomb {talk / contribs / physics / books} 16:53, 29 October 2016 (UTC)[reply]

Standards and related WikiProjects

It would probably be good to encourage and work with WikiProjects under whose jurisdiction certain identifier standards fall. For example, WikiProject Countries for ISO 3166, WikiProject Linguistics for ISO 639, WikiProject Writing systems for ISO 15924, and WikiProject Numismatics for ISO 4217. These sorts of things are not always on their radars. Thoughts? —Gordon P. Hemsley→✉ 07:06, 26 November 2016 (UTC)[reply]

Does anybody know what's going on with VIAF?

Every link, like VIAF Entry for William Shakespeare just ends up with a 404. Greetings from German Wikipedia. --Wurgl (talk) 11:17, 19 March 2017 (UTC)[reply]

VIAF seem to work on sundays too, it seems to be fixed. --Wurgl (talk) 16:19, 19 March 2017 (UTC)[reply]

Couple of requests

There are a couple requests at Template talk:Authority control. Sadads (talk) 14:24, 20 April 2017 (UTC)[reply]

A new newsletter directory is out!

A new Newsletter directory has been created to replace the old, out-of-date one. If your WikiProject and its taskforces have newsletters (even inactive ones), or if you know of a missing newsletter (including from sister projects like WikiSpecies), please include it in the directory! The template can be a bit tricky, so if you need help, just post the newsletter on the template's talk page and someone will add it for you.

– Sent on behalf of Headbomb. 03:11, 11 April 2019 (UTC)[reply]

ISBN

I am planning on doing another ISBN clean up run in a couple of weeks or months, unless there are any objections. All the best: Rich Farmbrough (the apparently calm and reasonable) 17:28, 10 May 2020 (UTC).[reply]