Wikidata:Project chat

On this page, old discussions are archived after 7 days. An overview of all archives can be found at this page's archive index. The current archive is located at 2026/02.

SpBot archives all sections tagged with {{Section resolved|1=~~~~}} after 1 day and sections whose oldest comment is older than 7 days.

There is a website EverybodyWiki (Q108492292) and I think it might make a good external identifier. But it is blocked from being linked to on wikipedias. Is there a reason we are avoiding it as an identifier such as malware? Immanuelle (talk) 14:52, 20 January 2026 (UTC)Reply

It's probably because the site is frequently used as a source for spam items and pages while it does not make the subject notable. Ternera (talk) 15:10, 20 January 2026 (UTC)Reply

The site basically collects articles marked for deletion and declined drafts on Wikipedia via a bot. It often contains spam and low-quality articles. Samoasambia ✎ 16:00, 20 January 2026 (UTC)Reply

On the other hand, it is still used, see for example Q136531364#P2888, Q113847934#P27. By defining it as an identifier, you could ensure that it is only used as a main value (and not as a reference) and the presence of such a property on an item gives an indication of notability for that Wikidata entry. Difool (talk) 02:59, 21 January 2026 (UTC)Reply

A untrustworthy site cannot give notability or be used as a reference. Adding properties for untrustworthy sites gives them legitimacy we do not think they should have, and would allow their proponents to insert links from WD we do not think should exist. Vicarage (talk) 07:45, 21 January 2026 (UTC)Reply

Then maybe Wikidata item for a source that does not imply notability (Q130719186) or something stronger should be linked to EverybodyWiki (Q108492292), so bots/tools can recognize that it is not valid to use as a reference. Difool (talk) 09:13, 21 January 2026 (UTC)Reply

Notability is not the same as reliability; there are also Wikidata property that suggests a reference is weak (Q113558322) (but only for properties) and source known to be unreliable (Q22979588). Q130719186, with the statements it currently has, seems to be an incorrect value for P31 where it is used, and makes WHOIS (Q590967) and Genealogics (Q19847326) Wikimedia internal items. Peter James (talk) 20:12, 21 January 2026 (UTC)Reply

I see that descendants of source known to be unreliable (Q22979588) are sometimes used, for example on The Guardian (Q11148): Q11148#P9852 and Q11148#P1343. It feels a bit convoluted maybe because Q11148 is not a Wikidata item.

Maybe it is an idea to add EverybodyWiki (or for example grokipedia.com, see Q4720205#P106) as a property and then explicitly indicate/enforce that it should not be used. This could be done with an empty property scope constraint (Q53869507) as is done now with Sandbox-Item (P369). Difool (talk) 03:50, 24 January 2026 (UTC)Reply

@Difool The example of where it used, are bad uses. I removed them. You should not add exact matches to websites that aren't really notable. You also should not use it as references. ChristianKl ❪✉❫ 15:38, 3 February 2026 (UTC)Reply

@ChristianKl Yes, I agree with you. I was just thinking about how you automate tasks that you just did manually. For both tools/bots and people, it is useful if the information "This website is not notable and should not be used as a reference" is stored somewhere in Wikidata, and that information is more easily stored in P## items than in Q## items. Difool (talk) 01:14, 4 February 2026 (UTC)Reply

Blocklists are the way to note bad sources, not properties. In this case, the core issue seems to be about trying to put much more than references, P973 and P2888 claims than is standard. ChristianKl ❪✉❫ 02:10, 4 February 2026 (UTC)Reply

No, there is no good reason for avoiding it; I strongly support unblocking it on English Wikipedia; and I support an external identifier property for everybodywiki.

That site has editable deleted Wikipedia articles which are often very useful and sometimes/rarely deleted against Wikipedia policy / without sufficient rationale. When an article does not or not yet belong onto English Wikipedia, it can still be useful, including for subjects that have Wikidata items. Please let me know if you propose such a property, thanks. Prototyperspective (talk) 12:32, 24 January 2026 (UTC)Reply

they say they actively encourage articles on anything, with no constraints at all, and take no responsibility for content. That seems a very good reason to avoid it. Vicarage (talk) 05:14, 26 January 2026 (UTC)Reply

That's a reason to not use low-quality articles and articles on unnotable subjects. However, there's also lots of articles on notable subjects including many deleted Wikipedia articles and the most comprehensive encyclopedic and/or wiki articles on notable subjects that didn't meet Wikipedia notability criteria because too few reliable sources reported on the otherwise notable subject. It also contains the proper articles for Wikipedia articles that were trimmed to useless length where it was previously the most comprehensive resource on the subject on the World Wide Web. If a property existed that doesn't mean it should be added wherever possible and it would only be set on Wikidata items so articles on EBW on other subjects that aren't notable wouldn't be set because of that either. Prototyperspective (talk) 12:21, 26 January 2026 (UTC)Reply

So you support strictly limiting the use of the Wiki and not just allow people to add it indiscriminately on every Wikidata item with an EverybodyWiki article? Trade (talk) 23:24, 28 January 2026 (UTC)Reply

These should only be added to existing Wikidata items which implies the subject is sufficiently notable for the page to probably be useful. Additionally, it doesn't make sense (it's not useful) to add EBW links where these don't provide any additional value such as when the article also exists on Wikipedia and not in a severely shortened version compared to EBW – for example when the EBW article is nearly the same or just a very old version without any usefulness (example).

.

Whether and how much to further limit the use is a question not as simple but I would generally support limiting it to just cases where the EBW article is useful and of sufficiently good quality (in this context, note that we even link wikia/fandom pages and often including some that are without refs and very short and not useful). Basically, people should only add the EBW link when it's useful and if it's added indiscriminately it won't be very useful anymore except for items that have nearly no other link or interwiki statements set. Prototyperspective (talk) 23:52, 28 January 2026 (UTC)Reply

Hello all,

See the case of William Walker Atkinson (Q1335019): American author. They have 4 pseudonyms. As far as I can see the practice here is, correct me if I'm wrong, All should be in the same item. So I gather all relevant VIAF's. As you can see it is too much. It would be good idea to update the practice and start creating items for pseudonyms. I men:
a. instance of (P31)pseudonym (Q61002) or instance of (P31)pen name (Q127843)
b. allowing links to pseudonyms in pen name (Q127843): pseudonym adopted by an author (or a group of authors) and printed on the title page in place of their real name from the original humen author.
@Wurgl (We talked about it a little bit.). Geagea (talk) 10:26, 21 January 2026 (UTC)Reply

Sounds good for me, what does Kolja21 say? --Wurgl (talk) 10:50, 21 January 2026 (UTC)Reply

Why is having 15 VIAF IDs "too much"? It is what it is, and it is temporary given that some of these IDs will (probably) eventually be merged by VIAF on their end. One could add subject named as (P1810) to identify which ID belonged to which pseudonym. Dogfennydd (talk) 12:36, 21 January 2026 (UTC)Reply

At least the National Library of Israel creates pseudonymous items as part of its normal work process. I believe that it is the case for other national libraries. I suppose that it is guidance taken in the field of librarians. Those VIAF's will not merged. What I'm suggesting that we behave the same.

Also, In the normal work process of librarians they are creating items based on books, so if the author is pseudonym, it's probably that they are creating items to pseudonyms. So I'm suggesting to create items to pseudonyms and connect them with the original author.

I'm sure that there are people hear from the field of librarianship that can understand what I'm talking about. Also, split the particular item above will made it neater. Geagea (talk) 14:25, 21 January 2026 (UTC)Reply

I'm a librarian and understand what you mean, but I think Wikidata is a very different kind of system than a library catalog. Library catalogs have multiple authority records for pseudonyms because they control names as they appear on works to create access points, rather than modeling the person as a single entity. Wikidata models people as single entities with multiple names. Listing pseudonyms on a single item makes sense in this context, even if it means long lists of external identifiers.

Changing the practice here would also be a big ask given that there are currently 100,000+ items that use the pseudonym (P742) property (compared to 598 that use it in instance of (P31)). There are certainly exceptions, like Robert Galbraith (Q110929251), but even that example is very messy with three properties that point back to J. K. Rowling (Q34660) in different, somewhat confusing ways. Mcampany (talk) 19:47, 21 January 2026 (UTC)Reply

I think WD dealing also with authority names and in many points there is a similarity of our working and library catalog. I don't think that change of attitude is a so big bother. We just need to choose the correct method of working. Robert Galbraith (Q110929251): pseudonym of J.K. Rowling is a good example. They are using three properties confusing ways because the issue not settled yet. We have similar topic with books and editions. We are allowing all editions and connect the editions and the books with from both sides. Geagea (talk) 21:26, 21 January 2026 (UTC)Reply

I agree that WD items about people and library catalog authority records are very similar, that's why I'm happy to contribute to WD as part of my library work. :) What I'm saying is that there already seems to be a widely-used method in WD for dealing with pseudonyms, even if it doesn't match the way we do it in libraries. It seems to work well for the most part, even if it means that there are sometimes messy items with 15 VIAF cluster ID (P214). That's part of reconciling two different data models, right? Creating items for the 100,000+ uses of pseudonym (P742) would have a big impact on people who already use that data, and I think that's why I'd be a little hesitant to do something like this. Mcampany (talk) 22:37, 21 January 2026 (UTC)Reply

My point about "messy" was not the main point. My point was that if there is 15 VIAF's in one item it will split to 5 items so each one have 3. And I think there's a place for items that the instance of (P31) is pseudonym. We are wiki not library and we are making changes all the time if it's better way. Geagea (talk) 22:57, 21 January 2026 (UTC)Reply

It’s incorrect to say that those VIAFs will not get merged; this happens regularly (and they often use Wikidata as a guide for IDs to be merged). There are three IDs which only mention Ramacharaka, four which only mention Dumont, and three which only mention Three Initiates, i.e. 7 redundant items from those three pseudonyms alone. Eight VIAF IDs for somebody with five different names is not excessive. Dogfennydd (talk) 23:49, 21 January 2026 (UTC)Reply

+1. We should not be guided by VIAF. The clusters change frequently. Multiple authority files for one item are perfectly acceptable if they are about pseudonyms. If you look at William Walker Atkinson some pseudonyms have their own GND, others are grouped together. Six separate items would result in chaos. --Kolja21 (talk) 00:29, 22 January 2026 (UTC)Reply

only the same names will merged because national libraries have more than one authority name for the human and the pseudonyms. Actually I suggested to act like library catalogs which is expressed in VIAF. I see that there is a resistance to my idea I give up. I don't think items about pseudonyms will make chaos just like the model of books and editions are not chaos. Geagea (talk) 07:12, 22 January 2026 (UTC)Reply

My point is that library catalogs do not necessarily create an authority file for every pseudonym. NKČR-AUT jx20060125014 (Theron Q. Dumont; pseudonym) and GND 107037130 (William Walker Atkinson, including Theron Q. Dumont) do not match. VIAF cannot cope with this contradiction. BTW: The opposite is true in the case of a quasi-synonym (Q2122467). In the case of topical terms one authoriy file can refer to multiple Wikidata items. --Kolja21 (talk) 07:53, 22 January 2026 (UTC)Reply

we have solution to that cases identifier shared with (P4070): qualifier, to be used on external identifier IDs, indicating another Wikidata item is also matched to this ID. or mapping relation type (P4390)narrow match (Q39893967) Geagea (talk) 08:09, 22 January 2026 (UTC)Reply

I didn't wrote "Wikidata cannot cope with this contradiction". I wrote about VIAF. In Wikidata we have indeed P4070 but the inflationary use of this property would make the maintaining of authority files more difficult. --Kolja21 (talk) 08:19, 22 January 2026 (UTC)Reply

"They have 4 pseudonyms" - Library of Congress has 7. Are there more in the Israeli library, if so, can you add them? How can it be stored in the current system if a pseudonym has no sex or gender? DataStructure26 (talk) 16:29, 22 January 2026 (UTC)Reply

@Geagea you added two more VIAF one containing an Israeli ID, which you didn't add - why? Especially since I asked for adding more Israeli. DataStructure26 (talk) 11:04, 23 January 2026 (UTC)Reply

probably missed. which exactly id's I have missed? Geagea (talk) 11:07, 23 January 2026 (UTC)Reply

The one that is in the VIAF cluster containing an Israeli ID. You can find the two VIAF that you added ("two more") in the edit history, which is conveniently linked from each item page, here https://www.wikidata.org/w/index.php?title=Q1335019&action=history . DataStructure26 (talk) 12:21, 24 January 2026 (UTC)Reply

BTW, this topic is not about the example but generally about pseudonyms. Geagea (talk) 11:09, 23 January 2026 (UTC)Reply

Of course, but you gave only one example which isn't in a good shape. There were 15(?) VIAF I added more now there are 25. There were four pseudonyms now there are 7. I also added data for LC (7 pseudonyms), ISNI (6 pseudonyms, 1 duplicate), Brazil (6 pseudonyms) and several more. Several are still incomplete, e.g. Israel, but I cannot read Hebrew, so I pointed a user to that gap, but no result yet. Some data can be seen in the VIAF clusters but the source databases are not reachable for me (e.g. Argentina, Australia). I also added variants for the Three Initiates pseudonym. There are many more gaps. Also the sorting could be improved, e.g. the id for the human should probably be first. DataStructure26 (talk) 13:32, 25 January 2026 (UTC)Reply

"I gather all relevant VIAF's" you now added two more https://www.wikidata.org/w/index.php?title=Q1335019&diff=2456993435&oldid=2456949980 DataStructure26 (talk) 11:01, 23 January 2026 (UTC)Reply

correction: I gather all relevant VIAF's that I have seen. Geagea (talk) 11:06, 23 January 2026 (UTC)Reply

It was easy to find 8 more. DataStructure26 (talk) 13:34, 25 January 2026 (UTC)Reply

The current system uses more storage space and doesn't represent the sources correctly:

1. duplicates information:
   1. pseudonyms and their variants stored in aliases
   2. pseudonyms stored using property pseudonym
   3. real name and their variants and pseudonyms and their variants stored using qualifier "subject named as" and "alternative name" for each external identifier
   4. sex or gender stored for each external reference if not the same as the human that used it
2. stores subject has role = pseudonym for each external record where the source says so
3. doesn't match external records which refer to the same pseudonym

DataStructure26 (talk) 16:09, 22 January 2026 (UTC)Reply

DataStructure26, Regarding your first comment: when I created this topic, there were only four pseudonyms her. Now I see there are seven. A pseudonym item should not include birth or death dates, but rather a start time (P580) - when it was first used. It may have a sex or gender (P21) (not necessarily the same as the human). It’s similar to type of fictional character (Q112265308), and each pseudonym may represent different character traits.

Regarding your first comment. it should not be duplications in the item. 1. pseudonyms and their variants stored in aliases. - that happened because it allows users to see the pseudonym name when writing it in the search bar before clicking search. But if my suggesestion accept it will be not allowed to add the pseudonyms as aliases of the humen. That is the meaning of authority name.

2. pseudonyms stored using property pseudonym - this is the correct place for them. not as aliases. I suggest that the pseudonym be linked to items of them. Geagea (talk) 01:00, 23 January 2026 (UTC)Reply

Just wondering. Richard Savage (Q123035003). Richard Savage is the pseudonym. The real name is Ivan Roe. Should the name of the item should be the human name or might be the pseudonym? Geagea (talk) 15:31, 27 January 2026 (UTC)Reply

How often does that happen and are there sources that support that? DataStructure26 (talk) 15:36, 27 January 2026 (UTC)Reply

no idea. The authority name of all id's in VIAF is Roe, Ivan. only wikidata Richard Savage Geagea (talk) 15:46, 27 January 2026 (UTC)Reply

In the item SBN has Richard Savage, SBN is from Italy. Review of item creation shows the creator is User:Marcok and has Italian as native language in the user page. No Wikipedia article linked to the item, so that cannot be a reason. DataStructure26 (talk) 15:59, 27 January 2026 (UTC)Reply

Labels are for people searching, so George Eliot and Cary Grant labels with Archibald Leach, Mary Evens as aliases, as every Wikipedia does. Explaining stage names and pseudonyms should be done with item properties. Having Grant and Eliot as minimal entries pointing elsewhere would confuse anyone who did not know, and irritate those who did. Vicarage (talk) 15:59, 27 January 2026 (UTC)Reply

Should instance of (P31) of George Eliot (Q131333) be pseudonym (Q61002) or human (Q5). Geagea (talk) 16:18, 27 January 2026 (UTC)Reply

Human. Vicarage (talk) 16:22, 27 January 2026 (UTC)Reply

+1. The main label depends on the focus: science-fiction author (Richard Savage) vs. non-genre novels (Ivan Roe). This is a case were also two items would be possible as he also appears to have been a philologist. (LCAuth n50049689: The breath of corruption, an interpretation of Dostoievsky, 1946 and A style of your own, a commonsense guide to clear English, 1972. Only one name given: Ivan Roe.) The identity seems clear, but a doubt remains. --Kolja21 (talk) 18:29, 27 January 2026 (UTC)Reply

PS: "No Wikipedia article linked to the item", that's true. But a 1955 novel by Richard Savage is linked: When the Moon Died (Q133557781). @Iamcarbon: FYI. --Kolja21 (talk) 18:51, 27 January 2026 (UTC)Reply

In this edition (1955) - author: Ivan Roe.

in this edition (1963) - author: Richard Savage. Geagea (talk) 19:44, 27 January 2026 (UTC)Reply

No, cover of the 1955 edition: Richard Savage. 2nd cover: Richard Savage. British Library: When the Moon died. By SAVAGE, Richard, pseud. [i.e. Ivan Roe] (1963). WorldCat is a tertiary source like VIAF. --Kolja21 (talk) 22:09, 27 January 2026 (UTC)Reply

Some external records store an ISNI, which can be used to solve the problem that in a Wikidata item that stores information about multiple identities belonging to one human the relationships between the pieces of information remain unclear when no qualifiers are used.

User:Geagea used Atkinson Q1335019 as an example above. I added the ISNI to several statements if the source provided the ISNI (https://www.wikidata.org/w/index.php?title=Q1335019&diff=2459052458&oldid=2459045482). Norway, Korea, Italy, France (idref, bnf) and Greece provide ISNI. Greece is the only one among these that has some pseudonyms without an ISNI.

Note that "subject named as" cannot be directly used for matching, as the names may differ even if the same script is used. The pseudonym "Three Initiates" has several translations, but the main form mostly is the English one.

Geagea - does this address some of your concerns? Note, the Israeli library doesn't provide the ISNI in their records.

DataStructure26 (talk) 19:40, 29 January 2026 (UTC)Reply

I didn't say I have a problem but suggested new approach to the issue of pseudonyms - to create new items to the pseudonyms, just the way the issue of editions of books handled. But my idea was not susseful and I give up. The section of "Pseudonym as label for item about human" made me think that I might be wrong. But maybe adding instance of (P31) human (Q5) and 61002) might sole it. Anyway, thanks for your eforts. Geagea (talk) 08:39, 30 January 2026 (UTC)Reply

Is there any mode to view what label had a deleted item? Gatto bianco (talk) 14:51, 21 January 2026 (UTC)Reply

Generally the label is shown in the deletion log. ArthurPSmith (talk) 21:32, 21 January 2026 (UTC)Reply

Q84262987 don't have. Gatto bianco (talk) 08:35, 22 January 2026 (UTC)Reply

@Gatto bianco: Since Q84262987 was recently deleted, it is still cached by search engines. A DuckDuckGo search for "Q84262987" (including the quotation marks) reveals this item was about Marie Abou Khaled, a Lebanese singer born 14 February 1995. Dexxor (talk) 09:11, 22 January 2026 (UTC)Reply

Q84262987: Marie Abou Khaled, Lebanese singer Bovlb (talk) 19:33, 22 January 2026 (UTC)Reply

@ArthurPSmith: I wouldn't say that's generally the case, without having checked all deleted data objects. However, there are several administrators who only specify the reason for deletion but not name a label of the data object. Perhaps it would be useful for the administrators to establish a standard procedure for this. --Gymnicus (talk) 21:54, 28 January 2026 (UTC)Reply

I'm currently writing Entitybase, a new scalable backend for Wikibase, and having now solved the scaling issues I started hacking on the 10y+ search problem that WMDE nor WMF has not been able to fix (my judgement).

The search problem is a hinder to the success of Abstract Wikipedia as was seen recently when @Denny Vrandečić (WMDE) had trouble finding a common item using the Wikidata search function on a live community session to use in a function. The current situation is unacceptable if you ask me.

So lets fix it!

I started by defining major subgraphs that can be used to filter search results in ElasticSearch (ES). See Wikidata:Subgraphs. The current iteration of this idea is to define major graphs manually there and generate WikiProject YAML based on SPARQL to filter away small projects.

The tags in the YAML can then be used to add labels to ES and this can be implemented in a decent frontend (Wikibase UI or other). WDYT? So9q (talk) 11:51, 24 January 2026 (UTC)Reply

As an aside this was written in the latest newsletter:

Dumps: We finished our research on dumps, especially looking into different ways to define subsets. We will publish the report with results by end of February.

Let's hope they create JSONL dumps if they are going to improve the dumps! So9q (talk) 11:56, 24 January 2026 (UTC)Reply

What is "ES"? Ainali (talk) 12:29, 24 January 2026 (UTC)Reply

ElasticSearch? Was mentioned before. DataStructure26 (talk) 13:05, 24 January 2026 (UTC)Reply

@So9q: The Wikidata: namespace should be reserved for pages that describe general guidelines, policy, things agreed by this community. This is a personal project of yours and anybody going to this page could be very confused about whether this is how current search in the Wikidata UI works. Please move this to somewhere under your User: space, or submit an RFC to get community consensus on whatever this is. ArthurPSmith (talk) 18:56, 24 January 2026 (UTC)Reply

Agreed. DataStructure26 (talk) 14:56, 27 January 2026 (UTC)Reply

Sounds reasonable. Would it be acceptable if we put a "draft" template on it like they do in Wikifunctions on similar pages? So9q (talk) 05:51, 28 January 2026 (UTC)Reply

It looks too specific. Such a page should start with a definition what a subgraph in Wikidata is or link to such a definition.

When the code starts with the class human it should be explained why. Similar to the other classes.

Then, why wikiprojects.yaml next? Why that property?

Why is the page named subgraphs and has

1. subgraphs
2. wikiprojects

?

Are there a subgraphs-subgraphs and wikiproject-subgraphs? If they relate to properties why not call them

1. type-subgraphs
2. maintainedBy-subgraphs
3. externalID-subgraphs (new proposal)

? DataStructure26 (talk) 18:15, 29 January 2026 (UTC)Reply

Thanks for the questions and suggestions. I'll move it to my user space now and perhaps create a new project that helps define useful slices of the graph aka subgraphs. So9q (talk) 07:35, 1 February 2026 (UTC)Reply

Is there a Mix'n'Match catalog that matches film items to YouTube videos (probably based on their titles)? It could be used to set trailers as well as especially links to the full film on YouTube where available.

I think both of these things could be very useful and one of the most tangible direct ways Wikidata could become useful in the real world outside of Wikipedia as there are items for a large fraction of films (not the case for other types of items; but also not the case for documentaries sadly).

If there currently is no such catalog, could it maybe be created?

Related ideas:

• Wikidata:List of documentaries available on YouTube (sortable table interface)
• Help talk:WikiFlix#WikiFlix for finding & watching free films on YouTube (Netflix-like interface)
• Wikidata:Bot requests#Request to import links/IDs of full films available for free in YouTube (2025-09-29)

Prototyperspective (talk) 12:46, 24 January 2026 (UTC)Reply

I think this could become the first large-scale use of Wikidata as people use it in some platform dedicated to distraction-free dedicated free films browsing (this isn't supported on YT itself for example) especially when it comes to short films and later also documentaries if we'd have more data on these. I don't know how to create a Mix'n'Match catalog, and don't know whether it's currently possible to match things based on the title like that (or what's needed to make it possible). Prototyperspective (talk) 14:57, 30 January 2026 (UTC)Reply

People thought the website was too new to justify Wikidata having an identifier property for it. Have this changed? Trade (talk) 01:19, 28 January 2026 (UTC)Reply

It was proposed for a property October 29, 2025, so people have had about 3 months to familiarize themselves with it now. There have been no shortage of opinions, but the one objective issue that I saw raised was that it might re-use the page names from english Wikipedia, if this is true then it obviates the need for a property. Can someone find a counter-example? Pages unique to Grokipedia in other words. Infrastruktur (talk) 15:24, 28 January 2026 (UTC)Reply

You seem to forget that not all Grokipedia articles have their own articles on Wikipedia. What about those? @Infrastruktur: --Trade (talk) 23:24, 28 January 2026 (UTC)Reply

That's what I was asking for examples of. I just found one: https://grokipedia.com/page/List_of_NoCopyrightSounds_artists . Seems to be more extensive than the list at en:NoCopyrightSounds. Naturally, if we have more like this, then that's an argument in favor of a new property. Infrastruktur (talk) 23:49, 28 January 2026 (UTC)Reply

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] Trade (talk) 00:24, 29 January 2026 (UTC)Reply

Guess you don't work in marketing, these aren't the examples I would have chosen, but the examples clearly show Grokipedia have their own articles which addresses the main concern of the last vote. I'm more interested in articles where the LLM might have enriched and added value to an existing notable thing. As I was the person to take consensus on the last vote, and I assume this is what you're fishing for, I don't have a problem with the proposal being re-opened at this time, thumbs up. Infrastruktur (talk) 20:22, 30 January 2026 (UTC)Reply

The interesting thing with Grokipedia, is that there's a lot of misinformation in the mainstream media about it. There are plenty of article about how Grokipedia has contains errors that don't contain any example of factually wrong claims and just claims that are politically incorrect opinion. The lack of criticism of where the website got factual issues wrong is remarkable.

When it comes to Wikidata, providing people who might want to compare Wikipedia articles to Grokipedia articles with an easy way to match them together is a value, that might be good for Wikidata to provide. This however does need someone doing the actual work of running the bot that fills the property with data. I would feel better about a property proposal if there would be a person who actually wants to do the work of filling the property with data. Given the way xAI (or now SpaceX) is run, I would not expect that they will do the work of feeding the data into Wikidata.

It's one aspect that they most valuable entries for a Grokipedia ID property are exactly those entries that you don't get by just trying to resolve all enwiki article titles and then add if the article resolves the relevant ID to Wikidata. ChristianKl ❪✉❫ 15:58, 3 February 2026 (UTC)Reply

There are three separate things that people put forward in this discussion and #Everybodywiki above:

1. Whether we want to provide cross-links of Wikidata items (and thus Wikipedia pages) and a 3rd party.
2. Whether a 3rd party indicates/provides notability.
3. Whether we allow referencing statements by a 3rd party (or: consider it a reliable source).

A property can exist even if only 1) is agreed on. We have Wikidata property for an identifier that does not imply notability (Q62589320) and Wikidata property that suggests a reference is weak (Q113558322) that can document the other two. --Matěj Suchánek (talk) 09:43, 4 February 2026 (UTC)Reply

Hello Wikidata community,

Please, I would like to request that the Idoma language (idu, ISO 639-3) be added as a supported language for monolingual text values on Wikidata.

Details: Language name: Idoma ISO 639-3 code: idu Wikidata item: Q35478 Writing system: Latin alphabet

Reason for request: Currently, it is not possible to add native labels, names, or descriptions in Idoma as monolingual text. Enabling idu would allow proper documentation of Idoma personal names, place names, works, and other data on Wikidata, improving coverage of African languages and supporting accurate representation of Idoma-speaking communities.

References: ISO 639-3 entry for Idoma: https://iso639-3.sil.org/code/idu Ethnologue entry for Idoma: https://www.ethnologue.com/language/idu Thank you for considering this request. I am happy to provide any additional information or references if needed. NatHaddan (talk) 14:58, 29 January 2026 (UTC)Reply

Hi, I'm sorry if this is not the right place for this. I'm having difficulty merging two pages.

shaft (shaft (Q2147038)) is the same as shurf (shurf (Q186570)). (I was trying to link the Wikipedia page for Shaft (civil engineering) with Шурф in Russian Wikipedia, and I think this was stopping me.) Slava570 (talk) 20:27, 29 January 2026 (UTC)Reply

They look the same to me too. I have merged them. Not sure why shaft (Q116678645) exists though. Will investigate. Vicarage (talk) 23:04, 29 January 2026 (UTC)Reply

Actually, now I've found mine shaft (Q556186) I'm not so sure, and have undone the merge. Vicarage (talk) 23:12, 29 January 2026 (UTC)Reply

according to shaft sinking "Shallow shafts, typically sunk for civil engineering projects, differ greatly in execution method from deep shafts, typically sunk for mining projects."

"Shallow shafts" here links to "Shaft (civil engineering)" but there is no separate English Wikipedia page for mine shaft. In the Russian site, I think "shaft (civil engineering)" should still link to шурф (Shurf). The Russian site says that a shurf is rarely deeper than 20-30 meters. Шахтный ствол would be mine shaft. Slava570 (talk) 23:38, 29 January 2026 (UTC)Reply

I think shurf (Q186570) should be "trial mineshaft", a subclass of mine shaft (Q556186), in turn a subclass of shaft (Q2147038), and shaft (Q116678645) should be merged with mine shaft (Q556186). Vicarage (talk) 10:05, 30 January 2026 (UTC)Reply

Merging Q116678645 with Q556186 would make the merged item and Q2147038 subclasses of each other. Subclasses of Q2147038 includes Q556186 and ventilation shaft (Q10565770), but not elevator shaft (Q760158), which is a subclass of Q116678645. Peter James (talk) 12:20, 30 January 2026 (UTC)Reply

We could have trial shaft shurf (Q186570) < mine shaft (Q556186) < shaft (Q2147038) < shaft (Q116678645), but I wish the last one had a better name and some actual content. And shurf (Q186570) seems to be mining only in some Wikipedias, and also archaeological in others, more the term "test pit" we don't have, as instead we have the ugly test pitting (Q123039189) technique. Vicarage (talk) 14:45, 30 January 2026 (UTC)Reply

I'm new to this and a bit confused about how this works, but if it helps, the Russian site seems to have more separate Wikipedia pages. The main heading is something like "manmade structures" (related to mining) (rough translations) which is divided into "underground manmade structures" and "open structures." "Underground manmade structures" is then divided into four categories: vertical, horizontal, diagonal, and other. Shurf is under "vertical" along with 5 others Slava570 (talk) 14:46, 30 January 2026 (UTC)Reply

A shaft is a (near) vertical engineered cavity through which things pass. Having 4 (and now 5 that Peter found) entries with 3 identical names seems bad, I can see the merit for quick 'n dirty mine shaft, an archaeological trial hole, a properly engineered mine shaft, and a building cavity, all under a general subclass of cavity. Vicarage (talk) 17:05, 30 January 2026 (UTC)Reply

Ok, so here's more info from the Russian site:

Can the general subclass be called "engineered cavity." That's roughly the same as the main heading there.

Next is "underground engineered cavity" versus "open engineered cavity" which includes karyer (quarry), transheya (trench), kanava (ditch), kolodets (well)

Under "underground engineered cavity" is "vertical cavity" (I think you're calling that a shaft)

Under that you have:

gezenk (winze)

furel (no link to English, a shaft that connects two adits, not sure the distinction between this and winze),

shahtnyy stvol (mine shaft, no page for this in English, links to Shaft sinking),

shurf (shaft (civil engineering); should it be renamed shallow shaft as per the shaft sinking page?),

slepoy stvol (no English link, defined as a vertical shaft with no aboveground exit, [literal translation is blind shaft, but I'm not sure if that's the same thing]),

rudospusk (ore pass)

If this is not helpful, please ignore. I only wanted to be able to connect shurf to shaft (civil engineering) in Wikipedia... Slava570 (talk) 18:55, 30 January 2026 (UTC)Reply

I think these two items (Q32335769 and Q49795485) could be merged, since both are about people born in the city of Pedreiras. Questionadora ávida (talk) 15:42, 30 January 2026 (UTC)Reply

If that is true, you would need to merge commons:Category:Births in Pedreiras (Maranhão) and commons:Category:People of Pedreiras (Maranhão) first — Martin (MSGJ · talk) 16:06, 30 January 2026 (UTC)Reply

 Not done They are not the same and should therefore not be merged. Births and associated persons are not the same. RVA2869 (talk) 16:09, 30 January 2026 (UTC)Reply

Q1957216#P1792 & Q1957216#P1464 RVA2869 (talk) 16:12, 30 January 2026 (UTC)Reply

Well, one of the items is called "Births in Pedreiras (Maranhão)" and the other is called "Category: People from Pedreiras, Maranhão". If you look at the Portuguese Wikipedia category linked to the first item, you'll see a list of people born in the city. If you look at the second category, you'll see the exact same people (such as Keila Simpson and Edmilson Costa), some more, some less, but still the same people. Furthermore, it's not a matter of being "associated with the city," but rather of having been born there. Questionadora ávida (talk) 16:39, 30 January 2026 (UTC)Reply

What if someone moves there at age 2, and subsequently becomes famous? They are "from" the place with their formative years, not the strict city of birth. Its why there are 2 lists Vicarage (talk) 16:55, 30 January 2026 (UTC)Reply

Okay, in that case, the merge isn't necessary? Questionadora ávida (talk) 17:03, 30 January 2026 (UTC)Reply

how to put more information in wikidata of the same article ~2026-67046-2 (talk) 05:28, 31 January 2026 (UTC)Reply

Can you be more specific? Which item? What information?

Anyway, I have put some introductory links on your talk page. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:07, 1 February 2026 (UTC)Reply

Why is it more efficient to store items with different attributes under one common article rather than creating separate articles for each item? B.Scalling (talk) 05:35, 31 January 2026 (UTC)Reply

@B.Scalling: is this a question about a Wikipedia project rather than Wikidata? If so, it would be better to ask at the Wikipedia project. If it is about Wikidata, do you have an example? TSventon (talk) 17:11, 31 January 2026 (UTC)Reply

By using {{#property:P856}} I can retrieve the link in P856 of a linked page here in Wikidata (in my case, this one). Language of work or name is a required qualifier for official website. How can I retrieve the value of "language" into Wikipedia? I mean: how can I ask "bring me the P407 of the P856 of this item"? CasteloBranco^msg 15:41, 31 January 2026 (UTC)Reply

You can use Module:Wd (Q24733825) for this. {{#invoke:Wd|property|qualifier|P856|P407}} -> https://portal.fpa.pt/seleccoes/selecao-a-masculina/ (Portuguese). {{#invoke:Wd|qualifier|P856|P407}} -> Portuguese retrieves the language only. To make the examples work here, I've manually specified the item as in {{#invoke:Wd|qualifier|Q786147|P856|P407}} but you don't need to do that if you use the module on the linked page. Now, I can't really explain exactly how this module works in a comment so I'd suggest you read the documentation. Warudo (talk) 14:06, 3 February 2026 (UTC)Reply

як мені зробити так щоб вікидані марка білого можна було додати в пошук щоб опублікувати що потрібно для цього зробити Степан Джумага (talk) 16:45, 31 January 2026 (UTC)Reply

I'm seeing the term "anthroponomy" appear in some publications ([13], [14], [15], [16], [17], [18]) and I want to create a Wikidata item for it, but I'm having a hard time really understanding what the term means (or if maybe there are several meanings using the same term). StarTrekker (talk) 01:50, 1 February 2026 (UTC)Reply

@SpBot, Revibot I: Why is this page archived by two bots, every day?? - Erik Baas (talk) 06:02, 1 February 2026 (UTC)Reply

@Erik Baas

SpBot only archives after the archiving template has been added and Revibot 1 after 7 days without responses. RVA2869 (talk) 11:29, 1 February 2026 (UTC)Reply

I am writing to respectfully request the restoration of the Wikidata item Q137901501 (Alireza Ghiasvand), which was recently deleted under the criterion of "Spam/advertising."

I would like to clarify that the purpose of this item is to facilitate Linked Open Data (LOD) for scholarly and professional identification, which is a standard practice within the Wikidata community for researchers and engineers. This item was created to centralize verifiable academic and technical identifiers, including:

• Academic Standing: I am a Master's student and researcher at Politecnico di Torino, specializing in Metal Additive Manufacturing and AI-driven process optimization.
• Verifiable Identifiers: The item includes my ORCID iD, Google Scholar ID, and ResearchGate profile, which are essential for academic disambiguation in the global knowledge graph.
• Professional Certifications: I hold internationally recognized NDT Level II certifications (VT, PT, MT) aligned with ASNT SNT-TC-1A and am an active member of the American Society of Mechanical Engineers (ASME).
• Scholarly Contributions: My work includes published research in areas such as "Attitude Control of a Satellite" and "Vehicle Impact Simulation", which are indexed in scientific repositories.

The information provided is strictly data-driven and factual, sourced from official institutional and professional records. It does not contain promotional language or intent. Restoring this item will ensure the integrity of the knowledge graph regarding my contributions to mechanical engineering and additive manufacturing research.

Thank you for your time and for the vital work you do in maintaining the quality of Wikidata. I am available to provide any further documentation required to verify these claims.

Best regards,

Alireza Ghiasvand Mechanical Engineer & Researcher Politecnico di Torino Alirezaghiasvand (talk) 10:48, 1 February 2026 (UTC)Reply

Please can you give the DOI of one or more of your peer-reviewed publications? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:04, 1 February 2026 (UTC)Reply

Can we please get people to raise these issues in the correct location, even if that means reverting changes here, else we just encourage them to clutter up this forum. Vicarage (talk) 19:26, 1 February 2026 (UTC)Reply

Written at the top of this page is "Wikidata project chat
A place to discuss any and all aspects of Wikidata: the project itself, policy and proposals, individual data items, technical issues, etc." Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:32, 2 February 2026 (UTC)Reply

this page is clearly not a place to discuss all of wikidata, its for general policy issues, not something like deletions that only administrators can change. For the rest of us its an unwelcome distraction. Vicarage (talk) 14:22, 2 February 2026 (UTC)Reply

I refer you to my previous comment. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:29, 2 February 2026 (UTC)Reply

For what it's worth, I appreciate seeing discussions like this. I considered adding myself in the past (based primarily on the existing RFC 8416: Simplified Local Internet Number Resource Management with the RPKI (SLURM) (Q56055023)author name string (P2093)D. Mandelberg), but decided against it after reading a few discussions like this. dseomn (talk) 19:03, 2 February 2026 (UTC)Reply

Agree. That may involve editing the text Pigsonthewing quoted. If there is a place to request undeletions, it should be easy to find and threads about such here be moved to there. Prototyperspective (talk) 12:50, 4 February 2026 (UTC)Reply

Here's your quick overview of what has been happening around Wikidata in the
week leading up to 2026-02-02. Missed the previous one? See issue #716.

Discussions

• New requests for permissions/Bot: Sapper-bot 2 - Task: Monthly updates of the Dead Sea level and the "lowest point" property for Israel, Jordan, and Asia.
• Closed request for permissions/Bot: LiteraryWorksMetaDataUploadBot - Task: Upload metadata for literary works from research project. (Approved)
• Other: Proposal to configure the Newcomer Dashboard for Wikidata

Events

• Upcoming events:
  • Call for Sessions: Wikimania 2026 is Coming to Paris! The submission form will remain open until March 1. Results will be published in May.
  • Who's hiding behind museum collections? - A Wikidata Edit-a-thon by WiNoDa Knowledge Lab (February 24th, 1:00pm-4:30pm CET). This online Edit-a-thon will focus on the many collectors that contributed to the Arachnida and Miryapoda collection of the Museum für Naturkunde Berlin. Join us to uncover the unspoken human networks that shaped the history of science! Register here: http://winoda.de/en/event/wikidata-edit-a-thon/

Press, articles, blog posts, videos

• Blogs: Strengthening the Latin American public domain with Wikidata
• Thesis: Hebeis, Maximilian (2026): Entity Matching for Person Records in Authority Files, Bamberg: Otto-Friedrich-Universität (Master thesis in computer science: "a case study into applying learningbased entity matching to person records from two large authority databases, namely the German national Integrated Authority File (GND) and the crowd-sourced open knowledge base Wikidata")
• Papers: Special Collection of the Journal of Open Humanities Data (JOHD): Wikidata across the humanities: datasets, methodologies, reuse, https://openhumanitiesdata.metajnl.com/collections/wikidata_across_the_humanities (new papers have been published during the last week; see a list at d:Q136325279#P527)
• Videos:
  • Africa Wiki Women 1st Newbies Community Meetup - held an orientation for new members on editing Wikidata, Wikipedia, and Commons, to help bridge knowledge gaps of African women on Wiki platforms.
  • Mapping Africa: Visualizing Knowledge and Communities - explores how African communities use maps and geospatial data to enhance Wikidata, improve Wikimedia articles, and make local knowledge more visible globally.

Tool of the week

• Wdlist : Create and manage a personal list of Wikidata items.

Other Noteworthy Stuff

• Wikimedia Deutschland Job vacancy - Architect of Storage Solutions - (freelance contract / 4-6 weeks)
• You can now nominate your favorite Wikidata research award for 2025. Call for Nominations now open until February 15, 2026.

Newest properties and property proposals to review

• Newest External identifiers: U.S. Copyright Office Public Records System name ID, Lenta.ru person ID
• New General datatypes property proposals to review:
  • value type of this mathematical property (type of the value of this mathematical property when applied to this object or this type of objects)
  • Gifu Prefectural Shrine Association ranking (One of the three rankings given by the Gifu Prefectural Shrine Association)
  • instances must not have (instances of this class or of its subclasses must not have given property or statement)
  • African Scientists Directory ID (identifier for a scientist in the African Scientists Directory)
• New External identifier property proposals to review: Center for Jewish History person ID, LINCS ID, Artdoc.Media film ID, Biographical Dictionary of British and Irish Architects 1800-1950 practice ID, Walloon-French dictionary ID, Russian Register of Film Distribution Certificates ID, Hentaigana for Academic Information Exchange ID, Identifiant inventaire Nouvelle-Aquitaine, Algemeen Nederlands Woordenboek ID, Woordenboek van Nieuwe Woorden ID, Instituut voor de Nederlandse Taal ID, Woordcombinaties ID, Etymologisch Woordenboek van het Nederlands ID, Database of the Southern Dutch Dialects concept ID, GiveSendGo ID, CrowdRepublic project ID, European Dictionary Portal ID, Gravestone Project Cemetery ID, Uzbek-Russian dictionary ID, Ruwiki ID, Swedish National Archives ID, GDUNO ID, JMRC person id, Catalunya Romànica ID, Mercantile Navy List ID, Igromania author ID, Danmarks Kirker, Medvik ID, Obálky knih ID, National Theatre in Prague artist ID, National Theatre in Prague performance ID, Alpine Linux Wiki article, Microsoft Store developer ID

You can comment on all open property proposals!

Did you know?

• Query examples: Item with a Wikipedia page in lang. X but not Y (source) - modify the wd: QID and the schema:isPartOf for your own purposes.
• Newest WikiProjects:
  • WikiProject Chess - Rubinstein-Memorial list
  • WikiProject Units of Measurement
  • Wikiproject Medicine - List of Italian Doctors
  • WikiProject Evanion Collection
• WikiProject Highlights: WikiProject every Politician / Canada - all Commons lists were updated.
• Newest database reports: Variety of reports for deletion candidates

Development

• Wikidata integration in the other Wikimedia projects:
  • We are making progress with identifying additional ways in which we can reduce the number of irrelevant changes from Wikidata showing up in recent changes and watchlist on Wikipedia and co.
  • We are discussing with pilot Wikipedias about turning on showing Wikidata changes by default in watchlist and recent changes (moving it from opt-in to opt-out)
• GraphQL: We are putting the last touches on a GraphQL endpoint for Wikidata. We are planning to get it live later this month for feedback.
• LDF endpoint: We are planning to decommission the endpoint in preparation for the migration away from Blazegraph on the 18th of February (phab:T415696)
• Mobile statement editing:
  • We are almost done with adding support for editing statements with quantity and monolingual text datatype and finished support for mathematical expression statements. We started support for editing coordinate statements.
  • We are continuing the work on showing constraint violations.
• Wikibase Ecosystem: We are in the process of setting up a demo system for ontology federation (i.e. using Wikidata Items as values in statements on another Wikibase instance)

You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.

Weekly Tasks

• Add labels, in your own language(s), for the new properties listed above.
• Contribute to the showcase Item and Lexeme above.
• Govdirectory weekly focus country:
• Summarize your WikiProject's ongoing activities in one or two sentences.
• Help translate or proofread the interface and documentation pages, in your own language!
• Help merge identical items across Wikimedia projects.
• Help write the next summary!

· Previous issue · Full report · Unsubscribe . JosefAnthony talk ·MediaWiki message delivery (talk) 15:41, 2 February 2026 (UTC)Reply

I've posted a query about tagging people in the text of the Bible with Wikidata Q-ids at en:Wikipedia:Reference_desk/Humanities#Existence_of_Wikidata-tagged_Bible_text. Namely, has anything like this already been done? I'm just posting a notification here because the Humanities Reference Desk might not be frequented by Wikidata experts who might be able to help. – BMacZero (🗩) 06:18, 3 February 2026 (UTC)Reply

There's https://en.wikipedia.org/wiki/Template:Wikidata_fallback_link that can be used for the purpose of tagging people within a Wikipedia text with Wikidata IDs. I'm not sure what the current Wikipedia norms are for when it should be used.

As far as bible text goes, just having the whole text in Wikipedia likely would not be welcome. Maybe, Wikiversity might host something like that? WikiSource might be another place, but I'm not sure how they think about annotated text. ChristianKl ❪✉❫ 16:07, 3 February 2026 (UTC)Reply

Thanks for looking! I should have clarified that this wouldn't be a Wikimedia-hosted project. I'll check out Wikisource to see if they're doing anything, though! – BMacZero (🗩) 18:03, 3 February 2026 (UTC)Reply

• Q1: say there is a state owned company bank Bank Rakyat Indonesia (Q623042), the shares are partly owned by org A (52,66% of the share), org B (0,53%), and the rest of the share are owned by the public (46,81%). I recently know that you can have qualifier of proportion but what item should I add for the "public (46,81%)"?
• Q2: are there any guidelines on whether one should use owner of (P1830) (owner of), or child organization or unit (P355) (child organization), or has part(s) (P527) (has part), or operator (P137) (operator)?

Thank you in advance FenyMufyd (talk) 04:29, 4 February 2026 (UTC)Reply

Adding ownership data to company items could be quite useful but so far I haven't seen such and I can't see it on the item you linked. Could you clarify, e.g. by linking an example item that has things set? Prototyperspective (talk) 12:48, 4 February 2026 (UTC)Reply

See Wikidata:WikiProject Companies/Properties. Jklamo (talk) 13:58, 4 February 2026 (UTC)Reply

These two items (Q109043435 and Q16888963) could be merged, since both discuss LGBTQ history in Sweden. Questionadora ávida (talk) 12:07, 4 February 2026 (UTC)Reply

 Done Ainali (talk) 14:06, 4 February 2026 (UTC)Reply

як мені зробити щоб дані про марка білого були в публічно як мені так опублікувати Степан Джумага (talk) 14:36, 4 February 2026 (UTC)Reply