close
Jump to content

Wikidata:Project chat

Edit links
Add topic
Shortcuts: WD:PC, WD:CHAT, WD:?
From Wikidata
Edit
BERJAYA
 SpBot archives all sections tagged with {{Section resolved|1=~~~~}} after 1 day and sections whose oldest comment is older than 7 days.

Systematic pattern of mass deletions using non-official criteria

[edit]
Latest comment: 3 days ago42 comments9 people in discussion

I report a documented pattern of mass deletions applying criteria absent from the official notability policy (WD:N), involving a coordinated chain between three editors.

During the period of July 1–15, 2026, Bovlb's deletion log records more than 500 items removed, concentrated via the Nuke tool. The operational flow is consistent:

  1. Bovlb distributes the unofficial template User:Bovlb/uw-notability, which imposes criteria more restrictive than WD:N.
  2. William Graham nominates items in bulk for deletion (RfD), using the essay Self-promotion as justification — an essay that is not official policy.
  3. Yamato Shiya executes the mass deletions via Nuke, without individual assessment of each item.

Documented evidence

[edit]
  • Absence of individual review — The deletion comment applied to Q139762140 (an organization) reads: "Person subject not notable...". The use of a boilerplate incompatible with the item type proves no case-by-case reading occurred.
  • Admitted divergence from policy — On the template talk page, Bovlb acknowledged that its content reflects "deletion practice" and not community policy, refusing to accommodate alternative interpretations.
  • Capture of the oversight channel — A report filed at WD:AN was closed by Bovlb within 26 minutes, without substantive resolution, evidencing that the local control mechanism is compromised.

Personal impact

[edit]

The following items of my authorship were deleted on July 8, 2026:

The same Graham → Yamato pair operated six days earlier against Vagan Arzumanian, confirming the pattern is not isolated to my case.

Request

[edit]

I request:

  1. Community review of the documented pattern, with participation of independent editors;
  2. Administrative intervention to suspend the application of non-official criteria until the notability RFC is concluded;
  3. Provisional restoration of deleted items for individual reassessment under current WD:N.

— Johnnytelles (talk) 01:53, 16 July 2026 (UTC)Reply

Can you please not create many different topics about ostensibly the same thing? Yirba (talk) 02:02, 16 July 2026 (UTC)Reply
I realise that you are new to the Wikidata community, but serious accusations of this kind should be accompanied by evidence.
This is, I believe, at least your sixth thread concerning the deletion of substantially the same group of items. Please keep related concerns in one discussion rather than repeatedly opening new threads.
Only a small fraction of the deletions I made during the period you identify used the Nuke tool. Your claim that my deletions were concentrated through Nuke is incorrect.
My template is part of an item-rescue workflow intended to help editors improve items before deletion is considered. It does not impose criteria more restrictive than WD:N. That claim is also incorrect.
I encourage you to focus on making constructive contributions to the project. Repeatedly posting unsupported allegations across multiple forums is not a productive use of community time. Bovlb (talk) 03:42, 16 July 2026 (UTC)Reply
In one of your many threads you have confirmed that one of them items is you. I presume the rest are connected to you. Secretlondon (talk) 10:19, 16 July 2026 (UTC)Reply
@Secretlondon Yes. One of the items is about me.

Rosa Parks was also the person on the bus. That didn't make her wrong.

Being the subject of a record does not invalidate the record.

  • Quis custodiet ipsos custodes?* The community. Together. Documenting what it sees.

That is exactly what I did.

— Johnnytelles Johnnytelles (talk) 13:08, 16 July 2026 (UTC)Reply

You think you're Rosa Parks? LOL. Looks much more like self promotion. We don't like spam. Secretlondon (talk) 16:02, 16 July 2026 (UTC)Reply
@Bovlb — Three responses to three claims.
On the Nuke tool: The deletion log records Q139762124, Q139790566, Q139762140 and Q140000366 deleted on 8 July 2026 at 08:31 UTC by Yamato Shiya with tag Nuke (log). Whether Nuke covered a fraction or all deletions in the period is beside the point: the criterion cited — "self-published" — does not appear in WD:N. The tool used does not change the criterion applied.
On the template: Four independent editors documented in Template talk that the template's criterion 2 is more restrictive than WD:N: Ainali (23 May, 24 Jun), QubeCube (26 Jun), TiagoLubiana (6 Jul), and Rodrigo Tetsuo Argenton (5 Jul). Your own response to those editors was: "I don't feel it's useful for this template to try to accommodate alternative interpretations of notability policy." That is not a rebuttal — it is confirmation that the template intentionally diverges from the policy.
On the number of threads: When a section is closed 26 minutes after opening — before other editors can respond — the only path forward is a new section. The proliferation you cite is a consequence of that closure pattern, not a separate violation.
One additional fact: The deletion reason applied to Q139762140 reads: "Person subject not notable." Q139762140 is a company, not a person. That single discrepancy is proof the items were not reviewed individually before deletion.
The same pattern was independently documented by Rodrigo Tetsuo Argenton on 6–7 July — six days before my items were deleted — involving the same editors and the same non-official criterion. Johnnytelles (talk) 22:30, 16 July 2026 (UTC)Reply
I did not delete any of those four items, so I'm not sure why you're tying these claims to my deletion log.
Rodrigo Tetsuo Argenton made similar allegations earlier this month, but they were never substantiated. When asked to explain them or provide evidence, he was unable to do so. Repeating those allegations here does not strengthen them. In any event, my item-rescue workflow is unrelated to the deletion of these four items.
You are, of course, welcome to request the undeletion of your items. I even wrote a guide explaining how to do so. What is not helpful is repeatedly opening substantially the same discussion in multiple venues while making serious but unsupported allegations. Bovlb (talk) 22:53, 16 July 2026 (UTC)Reply
@Bovlb A pattern is emerging in this discussion: you deflect, change the subject, close sections before others can respond, and say a great deal. But you do not respond to what is actually put to you. Two specific corrections below.
The log cited is Yamato Shiya's, not yours. I said so explicitly. The point is that "self-published" appears in Yamato's deletion reason, and that criterion does not exist in WD:N. Your original argument was about the Nuke tool fraction; my response was that the tool is irrelevant if the criterion is unofficial.
Regarding Rodrigo Tetsuo Argenton: four editors documented in Template talk that the template's criterion 2 diverges from WD:N. That record exists regardless of what happened in other threads.
One fact requires no allegation: the deletion reason applied to Q139762140, a company, reads Person subject not notable. That text is in the deletion log. Johnnytelles (talk) 00:15, 17 July 2026 (UTC)Reply
I looked at the template + archived request for deletion at Wikidata:Requests for deletions/Archive/2026/07/08#Bulk deletion request regarding User:Johnnytelles:
  • @1: Template:Uw-notability isn't more restrictive than WD:N. It restates WD:N's three criteria in plain language and flags self-promotion as a caution.
  • @2: William Graham's nomination explicitly cites WD:N: it begins with "Does not meet the notability policy."
  • @3: The archived bulk deletion closing message shows Yamato Shiya did review items individually: "Deleted all except Q139762399, merged into Q10387902."
Difool (talk) 01:46, 17 July 2026 (UTC)Reply
1. The template text says that that the item "must have", whereas N2 only says that an item "can be described" using sources. In fact, that N2 is so relaxed about the sources is a part of the ongoing RfC for the notability policy. Ainali (talk) 05:51, 17 July 2026 (UTC)Reply
I agree with Ainali. The wording in the template is stricter than WD:N, particularly regarding sources under N2. QubeCube (talk) 08:38, 17 July 2026 (UTC)Reply
@Ainali @QubeCube The distinction you identified is exactly the central point of the concrete case: "must have" versus "can be described".
The RfC on notability policy reform confirms the current policy. The proposal to make "self-promotion" a policy was put to the community. The answer was no. The current policy remains as is.
The template applied a criterion the community reaffirmed does not exist. The deletion of July 8 followed the template, not WD:N. That is on the record. Johnnytelles (talk) 12:04, 17 July 2026 (UTC)Reply
The deletion messages literally link to Wikidata:Notability. Yirba (talk) 12:14, 17 July 2026 (UTC)Reply
@Difool Looking at the same documents. What they say:
1. Template vs WD:N The template requires "both strong identifiers AND reliable sources". WD:N says "can be described using serious and publicly available references". The difference is not one of language: it is one of obligation. @Ainali and @QubeCube documented this in the same thread today. Bovlb responded to both on Template talk:Uw-notability: "I don't feel it's useful for this template to try to accommodate alternative interpretations of notability policy." The template's own author described it as a reflection of "deletion practice", not of policy. That is on the record.
2. Graham cited WD:N in the header Citing WD:N in the header and applying WD:N are different things. The criterion actually used in the nomination was "self-published". That criterion does not exist in WD:N.
3. Yamato reviewed items individually The deletion log of July 8, 08:31 UTC shows the Nuke tag: a mass deletion tool. One saved item does not prove individual review of the others. Concrete proof to the contrary: Q139762140 (MKPE) is a company. The recorded reason: "Person subject not notable." That error does not happen when the item is read.
The same Graham/Yamato pair, the same non-official criterion, appears in the case of Vagan Arzumanian six days earlier — @Vagan_Arzumanian. This is not an isolated case.
Against logs, there are no arguments.
Johnnytelles (talk) 16:38, 17 July 2026 (UTC)Reply
@Difool — I understand what you analyzed. I don't think we are looking at the same documents.

Let's compare. The template says: "items must have both identifiers and serious sources."

WD:N says: "can be described using serious and publicly available references." The word missing from WD:N is "both...and".

The template requires identifiers and sources. WD:N requires only sources. That is an additional restriction, not a simplification. "Self-promotion" also appears in the template as a caution criterion. It is not in any of the three points of WD:N.

This is not my reading: four independent editors — @Ainali, @QubeCube, @TiagoLubiana and @Rodrigo Tetsuo Argenton — documented exactly this on the Template talk page. Bovlb's response to them was: "I don't feel it's useful for this template to try to accommodate alternative interpretations of notability policy." That confirms the divergence.

William Graham's nomination header may cite WD:N. The criterion applied in the deletion was "self-published", which does not exist in WD:N. The header and the actual criterion are two different things.

On individual review: the deletion log for Q139762140, a company, records "Person subject not notable." That error does not happen when you read an item before deleting it.

The documents are available for any editor to verify. Perhaps a second reading would help.

Note: the same editor has opposed BulkPatrolBot, an independent patrol mechanism that would bring transparency to mass-deletion reviews. Johnnytelles (talk) 03:08, 17 July 2026 (UTC)Reply

Rather than starting from the assumption that a serious wrong has been done and then looking for evidence to support that conclusion, I encourage you to spend some time learning how Wikidata’s deletion processes and community norms actually work. You may still disagree with them, but your criticism will be much more persuasive if it is based on an accurate understanding of the project.
Doing that first would also be a more respectful use of the time and effort of the volunteers who are trying to help you. Bovlb (talk) 04:40, 17 July 2026 (UTC)Reply
You have already been asked to keep this complaint to a single discussion, and you currently have three open threads concerning substantially the same matter.
Please choose one thread to continue so that the other two can be closed. Bovlb (talk) 05:23, 17 July 2026 (UTC)Reply
@Bovlb Interesting timing to join this conversation. This was a dialogue with Difool.
The argument does not start from an assumption. It starts from three documents that anyone can read:
  1. Deletion log of July 8 [1]
  2. Template:Uw-notability [2]
  3. WD:N [3]
The comparison between the second and the third is what is being discussed. The second criterion of the template states that an item must have both identifiers and serious sources — but that requirement does not exist in WD:N. Four editors have already pointed this out recently on Template talk:Uw-notability: @Ainali, @QubeCube, @TiagoLubiana, @Rodrigo_Tetsuo_Argenton. The response to all four was: "I don't feel it's useful for this template to try to accommodate alternative interpretations of notability policy."
If the criterion exists in WD:N, cite where. If it does not, the issue is not about learning how Wikidata works. It is about whether a non-existent criterion was applied.
This is the fourth time you have responded to this thread without addressing the argument: "serious accusations should be accompanied by evidence" (July 16, 03:42), "I did not delete any of those four items" (July 16, 22:53), "spend some time learning" (July 17, 04:40), "keep this complaint to a single discussion" (July 17, 05:23). No logs. No policy citations. No facts.
The issue here is not about me learning how Wikidata works. It is about stopping to deflect the argument and applying the criterion that is actually in WD:N.
Johnnytelles (talk) 11:34, 17 July 2026 (UTC)Reply
It's true that the current notability policy says "can be described" and not "is described". However, if the item is not described using serious and publicly available references, then it is perfectly fair to ask for evidence that it can be described in that manner. If no evidence is put forward, then the item cannot be proved to be in compliance with the notability policy, and is subject to deletion. Perhaps it would be more precise if the template said that the subject of the item should have serious sources, rather than the item itself. But I think it gets the general point across. At the end of the day, if serious and publicly available references exist, then why not reference them? Yirba (talk) 12:11, 17 July 2026 (UTC)Reply
@1: The template never appears on your talk page: that page currently doesn't exist. Whatever the template's wording (and yes, there is a discussion on that talk page + a notability RfC), it didn't play a role in your case.
@3: The full deletion log for Q139762124 is "Does not meet the notability policy: Person subject not notable, although already own some publications, most of them were preprints and self published, the rest articles are not notable due to low readership and limited citation." Starts with WD:N, "self published" explains the WD:N judgment, it doesn't replace it. And a shared Nuke summary doesn't imply "without individual assessment.": Q139762399 was merged rather than deleted, a different outcome from the rest of the batch. Difool (talk) 08:05, 17 July 2026 (UTC)Reply

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ for reference

I have already opened a thread on the administrators' noticeboard. However, because Bovlb is an administrator, no action was taken. Predictably, some users claimed I was in the wrong; administrator noticeboards across the Wikimedia movement consistently protect sysops at all costs, even when their actions conflict with community consensus and policy.

"At the end of the day, if serious and publicly available references exist, then why not reference them?" - Yirba

What exactly is the expected reference for wallet (Q131740)? Not everything has an external reference, and evaluating this through a traditional Wikipedia lens is a mistake. In many instances, we are the reference. Take items like Foto '48 (Q131420151) or Salto a cavalo (Q53137546). We catalogue GLAM collections and act as their database. In this capacity, Wikidata functions as the secundary* source.

Why are they deleting entries of books, of books? Even if the books have ISBN they delete them. Who are they to decide what is notorious or not?

We are not Wikipedia. We do not just compile existing knowledge; we structure and sometimes create it. The only wiki of the Wikimedia Movement that needs a notability criterion is Wikipedia, because it needs a tertiary source; all the others, no. Rodrigo Tetsuo Argenton (talk) 15:59, 18 July 2026 (UTC)Reply

As per Wikidata:Introduction, we are a "secondary knowledge base". That means we record statements that exist in a primary knowledge base and provide a reference to the primary knowledge base. You might not like that, but it's simply how Wikidata works. Yirba (talk) 16:29, 18 July 2026 (UTC)Reply
"In many instances, we are the reference" >> and we shouldn't be the reference, we should organize references.
"Even if the books have ISBN they delete them. Who are they to decide what is notorious or not?" >> If with "they" you are pointing to admins, they are not. And being notorious is not the threshold for notability here. In previous parts of the discussion you presented yourself as the correct interpreter of our notability criteria, "notorious" however is never mentioned. It is true that notability criterion 2 leaves room for interpretation, but it is undisputed that serious and publicly available references (plural) are necessary.
Consistently across the Wikimedia movement some users who see their work deleted try to blow this up to an example of the huge and dishonest way the system of admins neglects community consensus. Luckily it is rather rare at Wikidata. Let's keep it that way and discuss matters, disagree at times, but not look for not existing conspiracies. --Lymantria (talk) 10:03, 20 July 2026 (UTC)Reply
Thank you both for your replies. It seems we need to clarify some fundamental concepts regarding data ontology and procedural governance on this project.
@Yirba:: I believe there is a fundamental misunderstanding of what a "secondary knowledge base" means in the context of linked open data. Being a secondary knowledge base does not mean every item must be backed by a tertiary encyclopedic article or an external website. It means Wikidata aggregates and structures metadata derived from primary sources.
When we catalog a GLAM collection, the physical artifact or the museum's internal registry is the primary source. When we catalog a published book, the book itself (along with its official ISBN registry) is the primary source. By structuring this metadata, Wikidata perfectly fulfills its role as a secondary aggregator. Imposing a Wikipedia-centric requirement on a structured database fundamentally misunderstands the semantic web.
@Lymantria:: I appreciate your linguistic correction regarding my use of "notorious" instead of "notable"-a classic false friend for speakers of Romance languages (notório). It is always refreshing when an administrator focuses on a minor vocabulary slip from a non-native speaker rather than addressing the core procedural issue at hand: the deletion of valid, verifiable bibliographic metadata.
Regarding your point that we need "references (plural)": an ISBN is not a random number. It is an internationally standardized identifier backed by national libraries, bibliographic agencies (like Bowker), and global distribution networks. A single published book inherently possesses multiple serious, publicly available references by the sheer nature of its legal publication and cataloging globally. Deleting books with ISBNs because of the identity of the user who typed the data is an indefensible data practice.
Furthermore, dismissing documented procedural concerns as a "conspiracy theory" is a convenient rhetorical device, but it does not change the facts. It is not a conspiracy to point out that an unapproved essay (Wikidata:Self-promotion) was used as the functional basis for expedited mass deletions without a standard RfD process. That is a documented procedural bypass.
When valid facts are deleted based on unapproved personal essays and linguistic pedantry, it is completely natural for the community to push back. I am not looking for conspiracies; I am simply asking that sysops follow the established consensus regarding what constitutes a factual database, rather than trying to apply Wikipedia's editorial biases to raw metadata.
Best regards,
Rodrigo Tetsuo Argenton (talk) 03:48, 21 July 2026 (UTC)Reply
ISBN-13 (P212) is explicitly tagged Wikidata property for an identifier that does not imply notability (Q62589320), so the claim that a book "inherently possesses multiple serious, publicly available references by the sheer nature of its legal publication" doesn't match how Wikidata treats it. An ISBN confirms a book was registered; it doesn't establish that the book has been described by serious references, which is what N2 asks for. Difool (talk) 04:17, 21 July 2026 (UTC)Reply
What makes this discussion confusing is that your items are quite different:
  • Rodrigo's items: a batch import of conservation areas. The main issue was that several lacked identifiers and had only Portuguese labels, making them hard to distinguish from nearby areas. Bovlb flagged this on 4 July, and since then Rodrigo and others have added sourcing, for example check: https://www.wikidata.org/w/index.php?title=Q140432476&action=history.
  • Johnnytelles items: a self-created cluster, himself, his company, and his work. These went straight to RfD and were deleted, citing WD:N.
Both sets of items are now being discussed together here as if they represent a single pattern.
IMO, a self-created cluster is a reasonable trigger to actually check N2; so not a different standard, just a reason to look rather than assume the identifiers are sufficient on their own. Difool (talk) 10:54, 21 July 2026 (UTC)Reply
@Difool @Rodrigo Tetsuo Argenton: The distinction you propose is not in WD:N. Johnnytelles (talk) 18:50, 21 July 2026 (UTC)Reply
No, but it's a useful heuristic. Yirba (talk) 19:00, 21 July 2026 (UTC)Reply
"Useful heuristic" — that phrase itself confirms what this thread has been arguing: it is not in WD:N.
The items were deleted citing WD:N. Not a heuristic. WD:N. At the time of deletion, the central item (Q139762124) had records issued by five independent institutions:
The other three items (Q139790566, Q139762140, Q140000366) are directly linked to this verified identity — and were deleted alongside it.
A heuristic absent from policy triggered deletions that cited policy. That is exactly what this thread documents. Johnnytelles (talk) 21:18, 21 July 2026 (UTC)Reply
@Difool, a direct question: is "self-created cluster" a rule or an opinion? If it is a rule, give me the link.
It is not in WD:N.
No identifier confers notability on its own. None. The IMDb ID does not. The DOI does not. The ORCID does not. Notability comes from the entity. A show performed by Elvis in an unknown venue does not make Elvis notable: he already is. But the show is a verifiable record of an established entity, and it belongs in Wikidata.
"Self-created cluster" is not in WD:N. Opinions do not delete items. Policy does.
Johnnytelles (talk) 00:02, 22 July 2026 (UTC)Reply
It's not a rule; if I came across a cluster it would be a reason to look closer, same as Yirba said. Difool (talk) 01:41, 22 July 2026 (UTC)Reply

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ I am genuinely a bit baffled by the direction of this thread.

I am not here to discuss my items. I am trying to discuss the broader posture of sysops deleting and flagging entries based on their own personal beliefs. I have no idea why my batch of Brazilian nature reserves (Unidades de Conservação) was even dragged into this as a justification...

Wait... sorry, did I read that right?

Are you actually saying that having labels "only in Portuguese" was the actual trigger for this scrutiny? Have I perhaps missed some unapproved personal essay dictating that local geographic data can no longer originate in its native language?

It does strike me as rather curious that this recent wave of warnings wasn't just sent to me, but to a remarkably large block of Brazilian editors. It takes a certain level of... confidence, shall we say, to approach contributors who have meticulously built 2,000, 5,000, or even 9,000 items, and mass-lecture them on how to create basic entries. I must confess, I wouldn't quite have the boldness to do that myself.

But to focus that level of systemic scrutiny on a group of a specific nationality, and then have someone explicitly confirm in this very thread that the underlying "problem" was indeed the use of the Portuguese language... well...

And what on Earth is the problem with information being added at a later stage, or by another volunteer? The way that was phrased made it sound as though making subsequent additions to an entry is somehow a bad thing. I swore it was literally in Wikidata's core definition that we are a collaborative environment. Isn't iterative improvement exactly how the Wikimedia movement is supposed to work?

Which brings me back to the actual issue I came here to discuss. Are we now officially swapping out community-approved rules for personal heuristics?

Because when Johnnytelles accurately pointed out that the criteria used to delete his well-sourced items are nowhere in the official policy (WD:N), Yirba explicitly replied: "No, but it's a useful heuristic."

"Heuristic". For anyone else who, much like myself, wasn't entirely familiar with such a complicated word, it turns out it is essentially just a fancy term for taking a "shortcut".

And if that wasn't enlightening enough, Difool has now generously chimed in to confirm this pattern. When Johnnytelles directly asked if the "self-created cluster" rationale used against him was a rule or an opinion, Difool candidly admitted: "It's not a rule [...] same as Yirba said."

So, let me get this straight: well-sourced items, backed by five independent national institutions (Web of Science, CNPq, FAPESP, Ministério do Trabalho, ISBN), were summarily deleted based on criteria that you yourselves are now openly admitting on this very page are "not rules"?

Why is it that certain sysops are allowed to take these procedural shortcuts and enforce unwritten opinions, whilst those of us who actually read the official policy and point out the contradictions are told we are in the wrong?

I just want to understand the current state of play. If the new standard is that individual shortcuts and opinions can freely override Wikidata:Notability, I would appreciate some clarity. Should we all start inventing our own personal shortcuts and enforcing them on others? Otherwise, I would respectfully ask that we stick to the rules as they are actually written.

Rodrigo Tetsuo Argenton (talk) 02:57, 22 July 2026 (UTC)Reply

I'm not a sysop, and I haven't spoken with any administrators about this. Johnnytelles asked for "community review... with independent editors," and that's exactly what I'm doing.
I never claimed Portuguese labels were the issue. My point was simply that the items I looked at had no identifiers at creation, so the label was the only way to tell them apart from nearby areas. If they had been Dutch labels (my own language), I would have said the same.
Here's how I see the sequence of events:
  1. Rodrigo imported items that were hard to identify.
  2. Bovlb raised concerns on Rodrigo's talk page.
  3. The items improved. Your items (Q140432473, Q140432476) were not deleted; they were improved by you and others over the following two weeks. That is good, I never suggested otherwise.
As for "self-created cluster”: it isn't a rule. It's simply a reason for me to look more closely. I made that clear when asked. It did not cause any deletions, and I never said it did. Difool (talk) 04:07, 22 July 2026 (UTC)Reply
@Difool: "self-created cluster" is not a rule. You said so yourself.
If a deletion requires a rule, and this is not a rule, then my deletion had no policy basis.
Yirba said the same about "significant publications": a useful heuristic, not policy. Two admins. Same conclusion.
My item was deleted without a rule to support it.
What happens now?

Johnnytelles (talk) 03:14, 22 July 2026 (UTC)Reply

@Rodrigo Tetsuo Argenton: The record is complete.
Two admins. Two admissions. "Useful heuristic." "Not a rule." Same conclusion.
Your pattern and mine are the same. The criterion was not in WD:N. It was applied anyway, as @Ainali, @QubeCube and @TiagoLubiana have already documented on Template talk.
In the RFU for JETSAR, I documented a case I found while researching the deletion log: an Indonesian academic journal, with no connection to me. Items with official identifiers recognized by the country's Ministry of Education, deleted under criteria absent from WD:N. Eight items. Three minutes. Individual review impossible.
The only thing my case has in common with that one is the pattern and the hand that deleted.
No argument is missing. What is missing is a reviewer. Without a conflict of interest.

Johnnytelles (talk) 03:37, 22 July 2026 (UTC)Reply

Difool and I are independent editors who looked at this and gave our opinion. I'm not sure what more you want other than to hear from someone who agrees with you. Maybe there's some miscommunication but I really don't know what more I can say other than to reiterate that the process is sound and the deletions are being made consistent with WD:N and other Wikidata policies. Yirba (talk) 07:41, 22 July 2026 (UTC)Reply
@Yirba: thank you for the reply. But there is an inconsistency that needs to be resolved before moving forward.

Difool used "self-created cluster" as the criterion. When asked directly, he confirmed that this is not an official rule. Now you cite WD:N.

These are two different criteria. One of them does not hold.

If the criterion is WD:N, the question is simple: which clause covers items with official identifiers recognized by the country's Ministry of Education? Those items had exactly that. Verifiable identifiers, issued by a public authority.

I am not asking for agreement. I am asking for the policy basis. Johnnytelles (talk) 09:46, 22 July 2026 (UTC)Reply

@Johnnytelles: The way I see it, verifiable identifiers issued by a public authority aren't enough to qualify as notable. For example, countries often allocate an ID to every registered company. That doesn't mean that every company is notable. It's true that WD:N doesn't specifically say which identifiers suggest notability and which do not. It's largely something determined by community consensus. Unfortunately, there isn't a clear-cut way to determine if a subject is notable or not. There will always be ambiguity. However, there is an unwritten standard that is used by the community, and items are judged against that standard. Yirba (talk) 10:26, 22 July 2026 (UTC)Reply
Yirba, could you explain why we should not have all registered companies in the world represented here, provided they can be reliably identified and described?
What I see here is a Global North standard being applied to the Global South as though it were universal.
Again, this is not Wikipedia. We should not be excluding items merely because someone decides, based on an unwritten interpretation of WD:N, that a subject is not sufficiently "notable". That becomes particularly problematic on a multilingual and multicultural project.
How is someone in London supposed to determine whether a local organisation in Pirapora do Bom Jesus is "notable enough"? How is someone unfamiliar with Swahili-language publishing supposed to decide whether a particular book deserves an item? The answer cannot simply be an unwritten standard developed largely from the perspective of editors who have access to the sources, institutions and forms of recognition common in their own part of the world.
And using online presence as a proxy for importance is even more problematic. Many subjects were historically important but never had that importance transferred into the digital world. Others exist in places where institutions, archives and local media are poorly digitised, difficult to search, or not indexed by the platforms most editors use. Their lack of an online footprint does not mean they lack relevance or verifiability.
This is undoubtedly where structural bias enters the project. People editing from relatively privileged information environments can easily mistake the visibility of a subject for its importance. A company, book, organisation or cultural institution from Europe or North America is much more likely to leave the kind of digital trail that editors recognise as evidence. The same standard, applied mechanically elsewhere, systematically excludes subjects from underrepresented regions and languages.
So my concern is not simply that WD:N contains ambiguity. It is that this ambiguity is being filled by an unwritten standard that is neither culturally neutral nor equally applicable across the world.
If an entity can be reliably identified through authoritative identifiers and described with verifiable statements, I think the burden should be on explaining what harm its presence in Wikidata causes, rather than starting from the assumption that it must prove some vaguely defined level of "notability" before it is allowed to exist here. Rodrigo Tetsuo Argenton (talk) 03:43, 23 July 2026 (UTC)Reply
@Yirba: thank you for the reply. But this sentence reveals a criterion substitution that needs to be named.

WD:N N2 requires that the item "can be described using serious and publicly available references." What you are applying is different: a prominence threshold, expressed as "significant publications." These two criteria are not the same thing.

On the identifiers: the ISSN is not an administrative registration number. It is assigned by the ISSN International Centre, an intergovernmental body created exclusively to certify serial publications as intellectual works. GARUDA is not a registry. It is the official academic index of Indonesia's Ministry of Education. Together, they prove exactly what N2 requires: that the item can be described using serious and publicly available references.

The company analogy does not apply. A company registration number is an administrative identifier. An ISSN certifies intellectual content with international registration.

What you describe as an "unwritten standard" is, in practice, a criterion that replaces N2 with prominence. You confirmed that this standard is not documented. That means there is no policy basis for the deletion. I ask again: where is this criterion written?

The goal of a policy discussion is not to determine who is right. It is to verify whether policy was applied correctly. That requires willingness to acknowledge when a criterion does not exist, when an analogy does not hold, and when a standard is undocumented. I am not asking for agreement. I am asking for rigor. Johnnytelles (talk) 11:34, 22 July 2026 (UTC)Reply

It's exactly for reasons like this that the notability policy is undergoing review. The community tries its best to uphold the current policy, but there will inevitably be disagreement in how that's done. It seems like you have high standards when it comes to rigour, which is fair enough, but I don't think the current policy can be upheld to that standard. We can only try our best. Yirba (talk) 11:44, 22 July 2026 (UTC)Reply
[edit]
Latest comment: 5 days ago4 comments2 people in discussion

These three pages have information about exactly the same matter. I was trying to link them together but I'm getting an error and cannot finish the edit. The Russian version is proposed for deletion which is apparently a sign of the Russian-associated propaganda.

There are other languages in the list if you dive into Russian and Ukrainian versions.

It would be the right thing to link them all together to the English version.

Maybe someone can help to resolve this. Thank you. ~2026-40444-88 (talk) 01:58, 18 July 2026 (UTC)Reply

@~2026-40444-88: defection (Q332988) is about the act of defection. defector (Q101007738) is about the person doing the defecting. The two items cannot and should not be merged. — Huntster (t @ c) 02:23, 18 July 2026 (UTC)Reply
The sky is blue, the grass is green, the piss is yellow. ~2026-40444-88 (talk) 06:49, 21 July 2026 (UTC)Reply
Why are you trying to merge items https://www.wikidata.org/wiki/Q332988 https://www.wikidata.org/wiki/Q101007738 together?
Who proposed merging them?
Are you okay? ~2026-40444-88 (talk) 16:40, 21 July 2026 (UTC)Reply

VIAF confusion

[edit]
Latest comment: 5 days ago6 comments3 people in discussion

I saw that some identifiers for Thomas Dampier (Q7788834), Thomas Dampier Bishop of Ely 1748-1812 relate to N. Thomas, French engraver c. 1750 - c. 1812.

I have created N. Thomas (Q140611040) for the French engraver and moved some identifiers. I have also created c:N. Thomas on Commons.

I think the problem goes back to 2013, when Q7788834 was created for en:Thomas Dampier and linked to http://viaf.org/viaf/31061989, which seems to relate to N. Thomas, by VIAFbot. Then bots have added identifiers for N. Thomas and identifiers for N. Thomas have linked to Q7788834, e.g. https://data.cerl.org/thesaurus/cnp01326059 has information about N. Thomas, a link to Q7788834 and a picture of Thomas Dampier.

Is this a common problem and what more needs to be done? TSventon (talk) 13:28, 19 July 2026 (UTC)Reply

You might find Help:Conflation of two people helpful. Bovlb (talk) 17:36, 19 July 2026 (UTC)Reply
@Lydia Pintscher (WMDE) pointed me toward this as I was working on a tool for graph visualization outside of Wikidata.
I was going to provide some static graphs or something, but managed to deploy the tool this morning (very much a work in progress in terms of the UX etc) however I hope you might find it useful?
Basically you can put in the 2 Wikidata IDS you are looking at, and view how the tool tries to explore the graph outside of Wikidata.
When conflicts are found they are highlighted, and it will try to help you resolve them too.
https://lll.addshore.com/resolve?input=Q7788834 & https://lll.addshore.com/resolve?input=Q140611040
Prepare for bugs, and let me know what you think! ·addshore· talk to me! 08:20, 20 July 2026 (UTC)Reply
@Bovlb: yes, Help:Conflation of two people is helpful, I have created Thomas Dampier (Q140622594) for Thomas Dampier Bishop of Ely 1748-1812.
@Addshore: I think your tool is a bit advanced for me, most of my Wikidata experience is linking Wikipedia articles and categories to Wikidata and adding simple statements. My new Thomas Dampier seems to be mostly fine https://lll.addshore.com/resolve?input=Q140622594, but my new N. Thomas has some problems, perhaps some identifiers need to move back to the conflation page. Are you planning to write a manual? TSventon (talk) 15:42, 20 July 2026 (UTC)Reply
I took another pass at the UX of the tool last night, with any luck it will slowly make more sense and become easier to use.
Looking at https://lll.addshore.com/resolve?input=Q140611040, the main thing it highlights is that the links out there on web end up including 2x isni different numbers currently, 2 different wikidata ids and 2 different libary of congress ids if you follow them all.
A large part of this is due to the fact that Thomas Dampier (Q7788834) still has the IDs on it, and that I don't have any special handling for conflation (Q14946528), however that is not the only reason.
As far as I can tell the 2 sources below which you link to from N. Thomas (Q140611040) still in turn link back to the incorrect thing on Wikidata.
I tried to include just the graph of these nodes (during hop2 of exploration) File:(lll) linked data explorer Q140611040 21072026T0730Z.png
These links in turn reveal more connections back into concepts that likely should not be conected. ·addshore· talk to me! 07:36, 21 July 2026 (UTC)Reply
@Addshore: thank you for your advice, I have moved four conflicted identifiers from N. Thomas (Q140611040) to Thomas Dampier (Q7788834). I have also added more biographical information and Volume 33 of Thieme-Becker as a described by source. Hopefully the remaining identifiers are not conflicted.
I can't see any problems with Thomas Dampier (Q140622594), which probably means I am not looking hard enough.
Is it possible to get the identifiers to disentangle themselves? It seems odd that a French engraver and an English clergyman with different names get confused even if they did live at the same time. TSventon (talk) 15:04, 21 July 2026 (UTC)Reply

AUGMANITAI items created by User:AndreasEhstand

[edit]
Latest comment: 1 hour ago14 comments7 people in discussion

Hi, the user @AndreasEhstand has, at time of writing, created 5,556 items about himself, his research programme, and all manner of related concepts. Does anyone have thoughts on whether any of this is actually notable?

Let's pick an item he's created at random (and it really was random): Preference Drift Misalignment (Q139027330). Apparently it's a neologism. So I can just make up new terms and create Wikidata items for them? The description makes no sense and appears to be cut-off. The statements don't describe what this is at all. I don't know how a neologism can be subject to copyright, but apparently it has a Creative Commons licence. It's also apparently available on a Github-hosted website except the link doesn't work.

Is it just me who thinks this doesn't belong on Wikidata? Yirba (talk) 18:13, 19 July 2026 (UTC)Reply

Hi Yirba, thanks for flagging this — you're right that many of these entries are in poor shape. Several have truncated descriptions and at least one broken link, and that's on me. To be transparent: these are terms from my own research vocabulary project, so I have an obvious conflict of interest, and I understand the notability concern. I'm happy to cooperate: I can fix the descriptions and dead links, and I'm fine with removing the entries that don't meet Wikidata's notability standards — including nominating the weakest ones myself. Please point me to the right process, or feel free to batch-nominate; I won't contest reasonable removals. Sorry for the cleanup this creates. AndreasEhstand (talk) 18:33, 19 July 2026 (UTC)Reply
@Yirba This should be something for WD:AN. RVA2869 (talk) 18:44, 19 July 2026 (UTC)Reply
Thanks, though usually I post to the Admin Noticeboard if I see that our rules have been broken. In this case, I wasn't too sure, and especially due to the sheer number of items involved, I thought it would be useful to get community input first. Yirba (talk) 18:51, 19 July 2026 (UTC)Reply
BERJAYA Support removal from wikidata under WD:N and WD:NOT for original research, a quick glance shows many of the items are self-published articles and concepts from self-published articles. For @AndreasEhstand, please consider using your own wikibase instance instead, iirc Wikimedia Deutschland will even host a cloud instance for you for free. IntensionalLogician (talk) 20:37, 19 July 2026 (UTC)Reply


Thanks, everyone — and IntensionalLogician, the Wikibase Cloud pointer is much appreciated; I'll set up an instance and move the vocabulary there. As the sole creator, I request deletion of my non-notable coined-term items — the ones that are instance of (P31) neologism (Q130989) (neologism), roughly 5,500 — and I won't contest. To be safe for the batch, please keep only my genuine publications/research and real-world entities (most have DOIs, and they are of course open to review too): Q138522830, Q138522835, Q138634675, Q138773774, Q138970385, Q139042690, Q139042691, Q139042692, Q139042694, Q139042698, Q139042701, Q139042704, Q139042705, Q139042707, Q139042708, Q139042709, Q139270039, Q139270040, Q139387178, Q139491246, Q139491267, Q139491268, Q139491269, Q139491270, Q139491284, Q139491285, Q139491286, Q139491295, Q139549783, Q139690594, Q139690608, Q139700038, Q139700042, Q139700049, Q139700054, Q139848987, Q139889950, Q140147800, Q140554521, Q140558386. Everything else among my neologism items can be deleted. Happy to provide the full Q-ID list or take this to WD:AN in whatever format is easiest for an admin. Thanks again for the patient and constructive feedback. AndreasEhstand (talk) 23:10, 19 July 2026 (UTC)Reply
How does this happen? They all need to go. Secretlondon (talk) 08:39, 20 July 2026 (UTC)Reply
So how are we resolving this? I suggested a specific deletion list at User talk:AndreasEhstand#Notability. Bovlb (talk) 14:28, 26 July 2026 (UTC)Reply
1751 / 5556. I'm opposed to keeping the Zenodo DOI publications despite their identifier. -- IntensionalLogician (talk) 14:43, 26 July 2026 (UTC)Reply
Agreed. We should definitely delete the ones the creator suggested. But we should review all the items created. If they aren't notable, they should go. Zenodo DOI does not suggest notability. I realise it's a lot of items to go through. If I get a chance, I'll have a go at it. Yirba (talk) 15:01, 26 July 2026 (UTC)Reply
I've found that, of the 5556 items, 5521 of them are instances of neologism (Q130989) and should be deleted as requested by the item creator. That leaves 35 other items. I've started reviewing each one, and it seems clear that most should also be deleted. That said, Bundesvereinigung Mittelstand in Deutschland (Q140554521) is an example of an item that I'd argue actually does meet our notability criteria and should be kept (but could do with some improvement). Yirba (talk) 01:14, 27 July 2026 (UTC)Reply
Quick clarification first: the comment above from 15:06, 26 July signed "~2026-41725-01" was me — I replied from my phone and didn't notice I had been logged out. Apologies for the confusion.
On substance nothing changes: I fully support deleting the ~5,521 neologism items, and I won't contest the removal of the other research-programme items either. I would only ask that two items be judged on their independent, non-Zenodo sources rather than batch-deleted:
Andreas Ehstand (Q138634675) — covered by three independent press articles (Süddeutsche Zeitung, Münchner Merkur, ispo.com) plus a university project page; more can follow of course by my side and will follow soon.
Faszination Padel (Q139549783) — a book published by Meyer & Meyer (2024) and listed in the German National Library catalogue; my co-authorship is verifiable in the publisher's official reading sample (already cited on the item).
I am also cleaning up these items (removing dead links, trimming statements). If the community concludes these two don't meet WD:N either, I will accept that. Thanks @Yirba for the careful review. AndreasEhstand (talk) 02:10, 27 July 2026 (UTC)Reply
Thanks. To start with, I've requested deletion of the 5,521 neologisms at Wikidata:Requests for deletions#Bulk deletion request: neologisms coined by AndreasEhstand, although it seems something went wrong with the formatting for some reason. Yirba (talk) 02:16, 27 July 2026 (UTC)Reply
@Bovlb @IntensionalLogician
Thanks for following up and for the concrete deletion list on my talk page.
I fully support batch-deleting the non-notable neologism items (the large majority). As previously stated, I will not contest those removals.
Regarding the smaller set of items with DOIs / research-programme entities: I am fine with the community deciding case-by-case. If the consensus is that even the Zenodo-DOI ones do not meet WD:N, please delete them as well. I will move the entire vocabulary to a Wikibase Cloud instance and will not recreate the deleted items here.
Happy to provide any additional lists or help with the cleanup process in whatever format is most practical for the admins.
Thanks again for the constructive handling.
— AndreasEhstand ~2026-41725-01 (talk) 15:06, 26 July 2026 (UTC)Reply
[edit]
Latest comment: 6 days ago1 comment1 person in discussion

is there any way to sort search results at https://www.wikidata.org/w/index.php?search=&title=Special%3ASearch&profile=advanced&fulltext=1 by number of items of item কল্কি (talk) 14:26, 20 July 2026 (UTC)Reply

Wikidata weekly summary 741

[edit]
Latest comment: 6 days ago1 comment1 person in discussion
BERJAYA
Here's your quick overview of what has been happening around Wikidata in the
week leading up to 2026-07-20. Missed the previous one? See issue #740.

Discussions

Events

Press, articles, blog posts, videos

Tool of the week

Other Noteworthy Stuff

Newest properties and property proposals to review

You can comment on all open property proposals!

Development

  • Wikidata Query Service: We are working on adapting the Query Service UI to get it ready for the move to QLever
  • Mobile editing of statements: We are continuing to fix small issues found in testing
  • Lexemes: We are continuing the work on adding support for Lexemes to the REST API
  • Sustainability:
    • We are preparing the migration of some database tables on the client wikis that track the usage of Wikidata's data in articles on Wikipedia and co
    • We are further investigating how many changes from Wikidata that currently get sent to Wikipedia and co don't actually lead to a change in the article to be able to reduce these cases further

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

Dutch lexical data is inconsistent and features a lot of redundancy.

[edit]
Latest comment: 5 days ago1 comment1 person in discussion

This is an expansion of a comment I already made here Wikidata talk:Lexicographical data/Documentation/Languages/nl/modeling#c-Seraph3ne-20260720120600-Clarity Around Verb Forms? however, after looking further at the data, I feel that there are a large number of issues with all of the Dutch lexical data.

Starting with verbs as mentioned in my original comment, there is inconsistency around how certain things are labeled, but there are also certain things that are simply incorrect. Lexeme:L1459448 lists Q22716 incorrectly on many of the forms. I also think that there may be unnecessarily repeated data, for example, L1459448-F5 and L1459448-F6 are listed as Q77768790 and listing formal forms seems to be common practice despite there not actually being a specific formal form. Presumably they are listed as they are used with the formal pronoun 'u' however, these forms are not distinct from the second person jij, or the third person hij and so are already listed on the page. Verbs also seem to be tagged as instances of Q1340504, even though they are already given lexical category Q24905. I also think the listing of forms for the first person, second person, and third person, plural may be considered redundant as these are also consistently the same.

There are also issues with nouns, primarily I believe there are a number of duplicate pages based on grammatical gender, for example L1418654 and Lexeme:L1455300 represent the same lexeme, however the second lists Q499327 instead of the more common Q1775461, the use of the masculine gender for this lexeme is simply a dialectism and does not represent an entirely new lexeme, as such I think this should just be listed on the original lexeme. The problem is that this is not an isolated case Lexeme:L1458907, Lexeme:L1458908, and Lexeme:L1458909 refer to the same lexeme, with different gender properties. Some nouns are also marked as grammatical number Q110786, which seems unecessary.

Adjectives also list certain forms inconsistently Lexeme:L1419968 and Lexeme:L1458491, list the relative forms with different tags, and generally they lack accurately descriptive tags on many of the forms.

I think a more thouroughly written standard would help address many of these issues, and many of the lexicographical data projects for other languages do have more thourough standards. Seraph3ne (talk) 12:49, 21 July 2026 (UTC)Reply

Feedback on IORO Publications journal-article items (Ijter)

[edit]
Latest comment: 4 days ago4 comments3 people in discussion

Hello — I operate the Ijter account, which keeps Wikidata in sync with journal articles published by IORO Publications, a small open-access publisher. I was recently granted a one-year IP block exemption for this task and, as suggested there, I've made a handful of edits and would like feedback before running the full back-fill.

Sample article items just created: Balancing Development and Environment: A Doctrinal Analysis of Bangladesh’s EIA Legal Regime (Q140645736), Selling Arms, Waiving Rights: The Structural Failure of US Human Rights Law in Weapons Transfers (Q140645737), The Impact of Diplomacy on the Balance of Power in International Relations: Challenges and Opportunities (Q140645738).

Each article item carries: instance of scholarly article (Q13442814) (scholarly article), title (label + P1476), DOI (P356), published in (P1433) linked to the journal item, publication date (P577), and author name strings (P2093). The parent journal items (e.g. International Journal of Law, Politics and Governance (Q140451759)) carry title, ISSN (P236), and publisher.

Does this modelling look correct, and is there anything you'd change before I create the remaining ~260 article items? I'd especially welcome guidance on author handling (name strings vs. resolving to author items) and any missing properties. Thanks! Ijter (talk) 18:22, 21 July 2026 (UTC)Reply

Thanks for asking about this. I'm not a fan of the use of first line (P1922) for (what appears to be) the abstracts, I've never seen this used for any other scholarly articles. Also does not meet the English description guideline for capitalization. I also have to ask, is the address used for work available at URL (P953) going to be persistent? If the DOI directs to it anyway, this should probably be sufficient. -- IntensionalLogician (talk) 20:39, 22 July 2026 (UTC)Reply
I personally like the idea of including abstracts but notice that Wikidata:Property_proposal/abstract failed. The consensus of the discussion at Property_talk:P1922#Scope in 2016 seems seemed to support adding abstracts, but is at odds with the conclusion of Wikidata:Property_proposal/abstract in 2020. Curious what others think of this. -- IntensionalLogician (talk) 22:02, 22 July 2026 (UTC)Reply
Thanks for digging into the history — that lines up with what I found: no
dedicated abstract property, and P1922's scope for abstracts is genuinely
unsettled. Given that, I'll leave abstracts out for now rather than have a new
bot lean on a contested use, and I'm happy to revisit and back-fill them later
if a clearer consensus (or a proper property) emerges. I've applied the other
two fixes — dropped P953 and corrected the description capitalisation — and
re-synced the three samples if you'd like another look. ~2026-41068-22 (talk) 00:19, 23 July 2026 (UTC)Reply

Person who appears never to have existed

[edit]
Latest comment: 3 days ago5 comments3 people in discussion

en:Edward Spring, 1st Baron Askeaton was deleted from en-wiki in 2018 as a possible hoax (See deletion discussion).

His entry has since been removed from ThePeerage website, and marked as "may never have existed" on Wikitree [4] with all family connections removed.

Yet we still have an item for him, Edward Spring, 1st Baron Askeaton (Q16821927), with links to profiles for him on Geni.com and Prabook that are not (yet?) deprecated.

What is the best way forward here? Deletion? Or retention with some kind of appropriate P31 warning and heavy deprecation of other statements, to flag the apparent invention for anyone who finds him still on another site? Do we have any established best practice or policy?

And: how many other en-wiki articles may there be, that have been deleted there as hoaxes, but still have items here? Jheald (talk) 22:19, 21 July 2026 (UTC)Reply

I reckon it should either be deleted, or perhaps we change it to an instance of human whose existence is disputed (Q21070568). Yirba (talk) 07:50, 22 July 2026 (UTC)Reply
Wikipedia hoax (Q125168941) also exists, or is that too strong? Jheald (talk) 11:32, 22 July 2026 (UTC)Reply
Well, we don't know for a fact that it's a Wikipedia hoax. There's some evidence that it is, but it's not 100% conclusive. Yirba (talk) 11:34, 22 July 2026 (UTC)Reply
It was deleted on wikipedia as a "blatant hoax" and this has not been challenged since 2018, so good enough for me. I have marked as deprecated the claim that he is a human. Deletion would also be a good option — Martin (MSGJ · talk) 08:55, 23 July 2026 (UTC)Reply

Help panel question on Q140661932 (10:44, 22 July 2026)

[edit]
Latest comment: 2 days ago3 comments3 people in discussion

how to add the statement --Dine ALEMTO (talk) 10:44, 22 July 2026 (UTC)Reply

@Dine ALEMTO: Hi, please see Help:Statements#Adding statements, but also note that we discourage people creating items about themselves. Yirba (talk) 11:37, 22 July 2026 (UTC)Reply
@Dine ALEMTO What is your purpose for creating a Wikidata item and who or what gave you the idea to create it? William Graham (talk) 12:10, 24 July 2026 (UTC)Reply

Aguacapa Dam and Hydroelectric

[edit]
Latest comment: 1 day ago12 comments3 people in discussion

Both Aguacapa Hydroelectric Power Plant (Q4694461) and Aguacapa Dam (Q113956791) are separate items. As far as I understand in cases like Hoover Dam (Q172822), the object should be an instance of both dam and hydroelectric power station. This would help link the articles about the Hydroelectric and the Dam construction. Thanks!

Tian2992 (talk) 21:41, 22 July 2026 (UTC)Reply 

SELECT DISTINCT ?item ?itemLabel ?connectsLabel
:WHERE {
:  ?item wdt:P31 wd:Q15911738.
:  MINUS {?item wdt:P31/wdt:P279* wd:Q12323}
:  ?item wdt:P2789 ?connects.
:  SERVICE wikibase:label { bd:serviceParam wikibase:language "en". }
:}
:
Try it!
has 99 other plants with a connects with (P2789) to dams, so the issue is not uncommon. And 10 like your example use impounded by (P4792) in error. A glance at other examples suggests a muddle without a simple solution, I suggest you ask at Wikidata:WikiProject_Energy Vicarage (talk) 08:07, 23 July 2026 (UTC)Reply
A power station and a dam are two different things, so generally I would suggest they use different items. The property connects with (P2789) looks like a good solution. Otherwise part of (P361) is always useful — Martin (MSGJ · talk) 08:43, 23 July 2026 (UTC)Reply
99 are not also dams, but 1500 are, so I'd not go against the status quo. Vicarage (talk) 14:02, 23 July 2026 (UTC)Reply
I will confirm with the Wikiproject, but in my opinion merge is viable. Items like the Hoover Dam have all the data from both a power plant and a dam because it is both, they are inexorably linked, in purpose, function and physically. Also I'd contribute a pattern I've seen at least in ES and EN (and some degree RU) Wikipedia in that articles in English tend to focus on the dam while in Spanish they are on the Power. I base my opinion in that a hydroelectric power plant dam is indeed both things (a dam and a power plant) at once by design. Tian2992 (talk) 18:31, 23 July 2026 (UTC)Reply
I would agree with you, but tread cautiously. And if a particular language WP has 2 articles, we need to have 2 items here, which might trip you up. Vicarage (talk) 19:07, 23 July 2026 (UTC)Reply
Please do NOT merge them. They are different things, even both are commonly used together. There are dams which are not linked to power stations and there are hydroelectric power stations which don't use dams. Some properties relate to the dam (height, length, reservoir created, etc.) and some relate to the power station (power output, etc.) Both can have different build dates, operators, etc. so very poor ontology if merged — Martin (MSGJ · talk) 19:38, 23 July 2026 (UTC)Reply
Please review the Wikidata:WikiProject Energy/Review-1#Hydroelectricity ontology proposal on why it would be correct ontology to merge them the item being (P31) both dam and hydro power station. That proposal covers how to tag the operational data you mention. On the linked Wikiproject discussion they do concede there might be cases where both are not useful but for most cases in which the dam IS a power station, it would seem the best case to have just one 1 item Tian2992 (talk) 21:15, 23 July 2026 (UTC).Reply
Thanks - I have reviewed this page and the linked discussion. It seems that @Amadalvarez, @Rehman and also @ScottDavis reached an agreement that dams are better in separate items. I would draw your attention to Red Rock Dam (Q7304911) and Red Rock Hydro Plant (Q116718619). The former was started 1960 and completed 1969. The latter was started 2014 and completed 2021. We could represent these dates using applies to part (P518) but I would argue that keeping separate items in this case makes a lot more sense. In the 2019 discussion Scott also points out that coordinates of the dam and station can differ by a few kilometres. — Martin (MSGJ · talk) 08:32, 24 July 2026 (UTC)Reply
So I connected both items with connects replacing impounded. The Dam already has the is part of property to the Hydro plant. How can they get linked across multilingual articles? Tian2992 (talk) 22:16, 24 July 2026 (UTC)Reply
Probably the easiest way is to use redirects and connect these to each item, [5] [6] — Martin (MSGJ · talk) 07:50, 25 July 2026 (UTC)Reply
Thanks! Will follow through using this pattern. Tian2992 (talk) 20:49, 25 July 2026 (UTC)Reply

Description discussion at Talk:Q5367197

[edit]
Latest comment: 3 days ago2 comments2 people in discussion

There's an open discussion at Talk:Body & Brain (Q5367197) about whether the item description meets the neutrality guidance in Help:Description. I'm affiliated with the organization and have disclosed that in the thread, so I'd welcome an uninvolved editor taking a look. Sources are laid out there. Arnelque (talk) 21:45, 22 July 2026 (UTC)Reply

I think the entry currently underplays the religion side. Calling it a cult is wrong and is typical of some POV warriors on en:, but it isn't just Qi Gong either. Secretlondon (talk) 11:19, 23 July 2026 (UTC)Reply

Adopted children

[edit]
Latest comment: 2 days ago4 comments3 people in discussion

How do you state that X has adopted Y, and is an adoptive parent?

I didn't find it in the faq and couldn't find an applicable statement.

Thank you for any help. Lou Crazy (talk) 02:23, 23 July 2026 (UTC)Reply

I found one example at Q10565568#P1038. Bovlb (talk) 04:49, 23 July 2026 (UTC)Reply
Mizuno Tadakane (Q10565568)relative (P1038)Mizuno Tadatō (Q10565488)kinship to subject (P1039)adoptive parent (Q20746742) Bovlb (talk) 04:51, 23 July 2026 (UTC)Reply
Or Mizuno Tadakane (Q10565568)father (P22)Mizuno Tadatō (Q10565488)object of statement has role (P3831)adoptive father (Q61740757)? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 07:52, 24 July 2026 (UTC)Reply

Merge

[edit]
Latest comment: 3 days ago4 comments3 people in discussion

Hello, can someone merge Q1202761 and Q3422151? Thanks! ~2026-41011-36 (talk) 09:19, 23 July 2026 (UTC)Reply

Looks like a bad idea. 100 m course, freestyle swimming VS 200 m course, fly swimming. --Wolverène (talk) 10:15, 23 July 2026 (UTC)Reply
Sorry,
  • Q1202761 (world record progression 100 metres freestyle)
  • Q18434233 (record du monde de natation messieurs du 100 mètres nage libre)
Thanks! ~2026-41011-36 (talk) 12:26, 23 July 2026 (UTC)Reply
The first item covers men's and women's records, the second only men's, Q14566559 (record du monde de natation dames du 100 mètres nage libre) covers women's. TSventon (talk) 12:37, 23 July 2026 (UTC)Reply

References for properties inferred from other properties

[edit]
Latest comment: 3 days ago4 comments3 people in discussion

if property A is inferred from property B (which has references), should A be, identically referenced by duplicating B's references, or left unreferenced, or use some other method to explain that A is inferred from B? this doesnt seem to be covered in Help:Sources. RoyZuo (talk) 14:29, 23 July 2026 (UTC)Reply

I think you would use inferred from (P3452). Here's an example: Kebnepakteglaciären (Q10544604)continent (P30)Europe (Q46)inferred from (P3452)Sweden (Q34) Bovlb (talk) 15:40, 23 July 2026 (UTC)Reply
@RoyZuo: Here's an example of how I've gone about it before, but I'm not sure if there's a better way: Q110714132#Q110714132$C6526827-FF2D-4D9F-A6E0-08F62151EC0F Yirba (talk) 16:21, 23 July 2026 (UTC)Reply
There are, at time of writing, 60 subclasses of inferred from attribute of entity (Q105958742) with a uses property (P3176) statement. You can find them at: Special:Search/haswbstatement:P279=Q105958742 haswbstatement:P3176 Yirba (talk) 16:24, 23 July 2026 (UTC)Reply

Help panel question on Q140681421 (08:36, 24 July 2026)

[edit]
Latest comment: 2 days ago2 comments2 people in discussion

I want to add business details on my company page --YuvaViikas (talk) 08:36, 24 July 2026 (UTC)Reply

@YuvaViikas What is your purpose for creating a Wikidata item and who or what gave you the idea to create it? William Graham (talk) 12:09, 24 July 2026 (UTC)Reply

Merge request: Q140676305 and Q63445736

[edit]
Latest comment: 2 days ago2 comments2 people in discussion

Please review DAU 2026 (Q140676305) and Dau (Q63445736). The Russian sitelink on Q63445736 and the English and Italian sitelinks on Q140676305 appear to refer to the same film DAU directed by Ilya Khrzhanovsky. The German and Portuguese sitelinks on Q63445736 may not match this subject and should be checked. Autremoi (talk) 11:34, 24 July 2026 (UTC)Reply

@Autremoi: Hi, I think these are different films with the same name, that are both part of a wider project. Dau (Q63445736) refers to the 2019 film Dau described in this BFI article. DAU 2026 (Q140676305) refers to DAU 2026, an upcoming film described on the official website. Yirba (talk) 17:39, 24 July 2026 (UTC)Reply

Request for comment (the future of Abstract Wikipedia)

[edit]
Latest comment: 2 days ago1 comment1 person in discussion

You are invited to voice your opinions in a request for comment about the future of Abstract Wikipedia. Thank you! Kowal2701 (talk) 12:24, 24 July 2026 (UTC)Reply

How to convey that an item is comprised of multiple other things

[edit]
Latest comment: 2 days ago11 comments3 people in discussion

I am working with items of historical monuments of Romania, adding instance of (P31) to them. When the monument is a single structure, it is trivial to assign a P31. When it's a collection of multiple similar structures it's not hard either, e.g. the original label for Chilii (Q42748692) translates as "[multiple] monastic cells" from Romanian, which means I can add instance of (P31)group of structures or buildings (Q18247357)has part(s) (P527)monastic cell (Q1776533) (and another qualifier with quantity (P1114) if applicable).

Please guide me on how to best approach Incinta fortificată a bisericii evanghelice din Valea Viilor (Q43113621). As per the original label in Romanian, it is a group of structures or buildings (Q18247357) comprised of one courtyard (Q309250), four tower (Q12518), and multiple annex (Q566749). My guess is I should use instance of (P31)group of structures or buildings (Q18247357) without any qualifiers, and then below use has part(s) of the class (P2670) with three claims inside. Here is how I imagine it: Special:Permalink/2522104401#P2670. With this approach, I have no clue on how to show that there is more than one annex (Q566749) but we don't know the exact quantity. Let me know how to do that, and if this approach in general is correct. Gikü (talk) 13:53, 24 July 2026 (UTC)Reply

I think you should be using — Martin (MSGJ · talk) 14:04, 24 July 2026 (UTC)Reply
Along with Chilii (Q42748692)instance of (P31)group of structures or buildings (Q18247357)has part(s) (P527)monastic cell (Q1776533)? I was going to suggest that but was thinking my question is too long as it is. Gikü (talk) 14:11, 24 July 2026 (UTC)Reply
I'd have P31 of fortified church (Q1571674), I'd not claim it to be a group of structures or buildings (Q18247357) as pretty much all such buildings are collections of structural elements like that, and I'd reserve the term for when the buildings were actually dispersed across an area. Your use of has part(s) of the class (P2670) is fine, though so far we rarely do add so much detail, as there are so many to do! Its better if the labels are the building's name, not a description of its features, which you are coding in its properties anyway. Having both Romanian and English labels is excellent. Descriptions are best of the form "fortified church in town, region, Romania", which gives more context, but does not overwhelm. Vicarage (talk) 14:04, 24 July 2026 (UTC)Reply
@Vicarage: Thank you for your detailed reply! First of all, appreciate your clarification that group of structures or buildings (Q18247357) usually means "buildings [...] dispersed across an area". For Incinta fortificată a bisericii evanghelice din Valea Viilor (Q43113621), and about 80 other items like it, I can't add instance of (P31)fortified church (Q1571674) because the item is actually part of a fortified church, namely Valea Viilor fortified church (Q17029338). In this case, what P31 would you recommend if not group of structures or buildings (Q18247357)? Side note, after my P31 exercise I will add all the part of (P361) (and potentially has part(s) (P527) the other way around, too) connections where applicable.
For my use of has part(s) of the class (P2670) I have everything in an Excel, so adding all the details on Wikidata is not an issue. If you can, please recommend what to do with the has part(s) of the class (P2670)annex (Q566749) claim such that I relay there are multiple annex (Q566749) but I can't use quantity (P1114) to say exactly how many.
Your point in the second half of your message, about "the labels [...] building's name, not a description of its features", I hope it's cleared out by my explanation above of how this is not the actual fortified church but a part of it; and these labels come from the Romanian monuments registry, that's what I meant in my first message by "original label".
Thanks! Gikü (talk) 14:28, 24 July 2026 (UTC)Reply
I think you are having too many levels of detail, with a church having a courtyard with towers. I see that each has a different Romania LMI code (P1770) We have the same problem with the Historic England records that delight in having records at nested levels with some titles being the vague "courtyard with 3 towers and a curtain wall", which makes little sense out of context as a WD item. Or we have a master "Kenilworth Castle and gardens" item with a nested "Kenilworth Castle" and "Kenilworth Gardens", when WP and the rest of the world just has one. It was a heroic effort to get every database entry in WD, but really we should merge them. For your items I suggest you have Valea Viilor fortified church (Q17029338) with multiple Romania LMI code (P1770) with qualifiers, and assign the architectural features to that item. There are real building complexes like a Palace that might justify a multi-level approach, but I don't think a single church does. Vicarage (talk) 15:29, 24 July 2026 (UTC)Reply
I don't think it's a problem to mirror the registry items. Having one article in WP but multiple nested items in WD does not hurt anyone, as long as we mark them very clearly with heritage designation (P1435). Note that, currently, Romania LMI code (P1770) has a uniqueness constraint.
One more counterargument I see for merging such sets of items is you have to draw a line somewhere. It would definitely not make sense to merge Tomb of Géza Albach (1829–1900) (Q42862132) (a grave) into Házsongárd Cemetery (Q714902) (a cemetery).
To return to my original question, I'll go ahead with my original thought and hope that it's clear that if I didn't use a quantity (P1114) then it means the quantity is simply more than 0 and if it were 1 then quantity (P1114)1 would have been used. Gikü (talk) 15:53, 24 July 2026 (UTC)Reply
If you are keeping the registry items distinct, make sure they don't have overlapping P31 values, otherwise queries will get 2 items when only one is appropriate. building complex (Q1497364) would be better for the sub-entries, as that's more close-knit than group of structures or buildings (Q18247357) Vicarage (talk) 16:04, 24 July 2026 (UTC)Reply
Yes, they are not supposed to have overlapping P31 values but your remark made me realize that I am working on the list of items currently without P31 on Wikidata, and a senseful check is only possible if I verify the entire list of monuments. I will take care of it.
I will also make the switch to building complex (Q1497364) except for items that do not involve [exclusively] buildings, like Q43143864 ("isolated graves"). Awesome discussion, thanks! Gikü (talk) 17:05, 24 July 2026 (UTC)Reply

I have made a few edits which reflect my understand of how these properties work, such as:

I'm surprised that Romania LMI code (P1770) links to an article on rowiki - it really should link to a reliable source for that identifier — Martin (MSGJ · talk) 17:22, 24 July 2026 (UTC)Reply

Discussed here. monumenteromania.ro seems to be a good database but I don't think it's official, and I'm not sure if it's possible to easily build URLs to that site based on the LMI codes, e.g. Valea Viilor fortified church (Q17029338) vs [7]. Gikü (talk) 18:41, 24 July 2026 (UTC)Reply

Importing metadata of Leiden university dissertations collection to Wikidata

[edit]
Latest comment: 2 days ago2 comments2 people in discussion

I'm working with a group from Internet Archive. They have a collection of 162469 scanned dissertations from all over the world that were collected in Leiden University. They have the following relevant metadata fields:

collection theses-and-dissertations,leiden-university
contributor archive.org
creator the author of the dissertation
date the publication date of the dissertation, e.g. 1927-03-25T00:00:00Z
description I plan to completely ignore this. Only 123 of the 162469 dissertations have this - very miscellaneous text, e.g. "xvi, 112 p. ; 23 cm,Originally presented as the author's thesis (doctoral--Universität Zürich),Includes bibliographical references"
external-identifier Only 3158 have this - probably useful when it's there. It's a comma-separated list of external links to the item on IA in different formats, for example: "urn:lcp:IA41507204_0114:epub:1a027f61-bad7-4071-a56e-fdfebfcb7899,urn:lcp:IA41507204_0114:lcpdf:8b7601f3-ba8d-4241-8409-e9710a7b9b45"
foldoutcount The number of foldout pages in the book - it's an integer, mostly 0 but a bit more than a thousand have a number between 1 and 86.
identifier The Internet Archive ID, for example "IA41553408_0020"
imagecount That's the number of scanned images that the IA scan has.
language The language of the thesis, e.g. "French", "German", "English"
publicdate When it was published to IA - time/date value like 2026-06-05T03:30:15Z
publisher The company that published the scanned version of the thesis - often not the university!
title The title of the dissertation
volume the volume - can safely be ignored in the bulk import: there are two dissertations that were published in 3 volumes. The rest are empty, except for two which seem to have an erroneous value of 3767
year I think this is the year of the dissertation's submission.

I'd like to do a bulk upload. Is the best way of doing this to create a bot? Slashme (talk) 15:58, 24 July 2026 (UTC)Reply

Look at QuickStatements 3.0, but discuss on a relevant WikiProject first to check notability and get the properties correct, lots of them mention dissertations Vicarage (talk) 16:10, 24 July 2026 (UTC)Reply

strada regionale (Q16917915) and strada regionale (Q3975665)

[edit]
Latest comment: 1 day ago2 comments2 people in discussion

Same stuff? Bouzinac 💬●✒️●💛 05:40, 25 July 2026 (UTC)Reply

Looks like it. → ← Merged, thanks. Yirba (talk) 11:17, 25 July 2026 (UTC)Reply

Input wanted Linking OSM objects to Structured Data on Wikimedia Commons (MediaInfo entities):

[edit]
Latest comment: 1 day ago1 comment1 person in discussion

I have started a discussion in the OpenStreetMap community about improving the way OSM objects link to files on Wikimedia Commons:

"Linking OSM objects to Structured Data on Commons (MediaInfo entities)"

The idea is to explore whether OSM should be able to reference Commons MediaInfo entities (M-IDs) instead of (or in addition to) traditional file names. Today, an OSM object typically links to a Commons image using a filename: image=File:Example.jpg

However, filenames are human-readable labels that can change over time. The underlying MediaInfo entity (for example, M20159149) is the persistent identifier for the structured data associated with the file. Using the MediaInfo identifier would align better with Linked Data principles and the FAIR principle F1 (globally unique and persistent identifiers).

Would this make integration between OSM and Wikimedia Commons/Wikidata ecosystem easier?

The OpenStreetMap discussion is here: https://community.openstreetmap.org/t/linking-osm-objects-to-structured-data-on-commons-mediainfo-entities/145727

I hope we can get input from both communities before any concrete proposal is developed. Thanks!

- Salgo60 (talk) 12:11, 25 July 2026 (UTC)Reply

[edit]
Latest comment: 1 day ago6 comments3 people in discussion

Message: Hello, There appears to be an incorrect sitelink on Q2588 (28 May). The current sitelink is: rskwiki → Категория:28. май This is incorrect because it links a category instead of the corresponding article. The Rusyn Wikipedia article is: https://rsk.wikipedia.org/wiki/28._май The incorrect category sitelink prevents the article from being connected to the correct Wikidata item. Neither I nor an administrator of the Rusyn Wikipedia have been able to remove or replace this sitelink through the Wikidata interface. Could someone please remove the incorrect sitelink rskwiki → Категория:28. май and allow the article rskwiki → 28. май to be linked to Q2588? Thank you very much for your help. Saslavik (talk) 16:52, 25 July 2026 (UTC)Reply

✓ Done with this edit Kinsio (talk ★ contribs) 17:54, 25 July 2026 (UTC)Reply
Thank you very much. I have one question: why I am not able to correct it at wikidata, why I do not have the possibility to edit it. I had a problem with correcting wrong links before because I was not allowed to edit or delete them. Thank you. Saslavik (talk) 18:37, 25 July 2026 (UTC)Reply
I don't think there are any user rights issues involved, but it may be the same thing I ran into while doing the edit myself: I tried just pasting in the page title at first and wasn't able to submit the edit, but I eventually figured out that what I needed to do was delete one character to bring back the drop-down menu so I could choose the right page from there. If you haven't chosen an option from the drop-down menu, Wikibase won't let you submit the edit. Does that sound similar to what happened to you? Kinsio (talk ★ contribs) 18:50, 25 July 2026 (UTC)Reply
I don't have the Edit option for the sitelink table. Saslavik (talk) 19:17, 25 July 2026 (UTC)Reply
@Saslavik: Because it is widely utilised, the item is semi-protected so you need to be a confirmed user to edit it. Keep editing Wikidata and you'll be autoconfirmed. Yirba (talk) 19:49, 25 July 2026 (UTC)Reply

Qualifier or new properties for menu and online ordering URLs?

[edit]
Latest comment: 1 day ago6 comments4 people in discussion

Something I thought of a while back: Since it's OpenStreetMap (OSM) common practice to link restaurant chain locations to Wikidata for data about the entire chain, partly thanks to the Name Suggestion Index(62108705), Wikidata can be used to store properties that pertain to the chain in general rather than the particular location, such as the information that would otherwise be in the website:menu=* or website:orders=* OSM tags of a single-location restaurant.

What I'm wondering in regard to those two in particular is whether it would be appropriate to propose new properties for them, since there don't currently seem to be any directly equivalent Wikidata properties, or would something like URL (P2699)URLobject of statement has role (P3831)menu (Q658274)/URL (P2699)URLobject of statement has role (P3831)online ordering (Q115734960) be the best way to do it? Kinsio (talk ★ contribs) 17:17, 25 July 2026 (UTC)Reply

I think the qualifiers approach is best, because those use cases are pretty specialist. We don't want endless URL properties for all the things a web page can be. Vicarage (talk) 17:34, 25 July 2026 (UTC)Reply
I understand that argument, although we already have 123 URL properties, many of which are to record different types of web page. So a "menu URL" property wouldn't exactly be out of place. Yirba (talk) 20:13, 25 July 2026 (UTC)Reply
Most of those are external IDs. And inevitably someone argues that official shop URL (P10225) is not quite right for food ordering, or that a menu URL is needed as well as a pricelist_URL, Vicarage (talk) 20:45, 25 July 2026 (UTC)Reply
Rather than creating a new property would official shop URL (P10225) with a qualifier be sufficient? William Graham (talk) 20:30, 25 July 2026 (UTC)Reply
If you can actually order through that URL, then maybe. But if it's just a menu, then I don't think it would be appropriate. Yirba (talk) 20:44, 25 July 2026 (UTC)Reply

How to create Q-items for dictionary entries about words

[edit]
Latest comment: 1 day ago1 comment1 person in discussion

My main question here is how to connect these Q-items to Lexemes and/or Wiktionary. I'm not trying to recreate it, but take this as an example. There would be one main-space sub-page (on Wikisource) for each dictionary entry, and one Wikidata Q-item per sub-page. With encyclopedia articles, I use P921: main subject relative to it, but that doesn't work for lexemes. TE(æ)A,ea. (talk) 21:29, 25 July 2026 (UTC)Reply

WDQS not reflecting confirmed P50 statement (Q140686977)

[edit]
Latest comment: 14 hours ago3 comments2 people in discussion

I added a P50 (author) statement to Q140686977 via QuickStatements. This is confirmed successful via multiple independent checks: * The QuickStatements batch report showed the line succeeded. * The item's raw JSON at Special:EntityData/Q140686977.json shows a fully well-formed statement: datatype wikibase-item, value Q140619827, with a P1545 (series ordinal) qualifier of value "1" (string) — not a partial or malformed statement. * The item's rendered page displays the statement correctly. However, the following query, run repeatedly over 15+ minutes after the edit, still returns false: 

 ASK { wd:Q140686977 wdt:P50 wd:Q140619827 . }

I've ruled out client-side caching: tested in both Firefox and a fresh Chrome window, with hard refresh in both, and results were unchanged. The WDQS lag dashboard showed lag at approximately 45 seconds / near-zero throughout testing, so this isn't explained by normal update lag either. This is part of a batch of ~100 items imported via QuickStatements on 24 July 2026 (tag #temporary_batch_1784911052725). I'm concerned the same propagation issue may affect the rest of the batch, which would block linking these items' authors properly (P2093 → P50 conversion for Scholia author-profile purposes). Any insight into whether this is a known/current WDQS issue, or something specific to this batch, would be appreciated. Kriitters (talk) 10:55, 26 July 2026 (UTC)Reply

@Kriitters: Hi, please take note of the graph split. The query returns true with the scholarly graph: https://w.wiki/SnqJ Yirba (talk) 11:59, 26 July 2026 (UTC)Reply
Yes! Thanks SO much! Kriitters (talk) 13:15, 26 July 2026 (UTC)Reply

Help panel question on Q68050127 (12:08, 26 July 2026)

[edit]
Latest comment: 15 hours ago2 comments2 people in discussion

Муртазо --Муртазо (talk) 12:08, 26 July 2026 (UTC)Reply

@Муртазо: Hi, can I help? Yirba (talk) 12:12, 26 July 2026 (UTC)Reply

A multidimensional ontology for comparative political science

[edit]
Latest comment: 12 hours ago2 comments2 people in discussion

Over the past three weeks I have written and substantially expanded a series of articles on comparative political science for the Danish Wikipedia.

During that work I repeatedly encountered conceptual inconsistencies in the way related concepts are represented in Wikidata. This led me to undertake a systematic review of the existing ontology and to develop the proposal presented here.

The project consists of four documents:

  • Report
  • Proposed ontology
  • Bibliography
  • Request for Comment

The proposal does not suggest immediate changes to individual Wikidata items. Instead, it presents a conceptual framework intended as a basis for discussion before proposing specific implementations.

Project overview:

Request for Comment:

Comments and suggestions are very welcome.

IvarT (talk) 12:58, 26 July 2026 (UTC)Reply

I suggest you advertise it at Wikidata:WikiProject_Ontology where the experts reside. Vicarage (talk) 14:35, 26 July 2026 (UTC)Reply

Merge?

[edit]
Latest comment: 6 hours ago3 comments3 people in discussion

I think Michaela Mann (Q6835607) and Michaela Mann (Q95698379) are about the same person. Dugnad (talk) 16:54, 26 July 2026 (UTC)Reply

→ ← Merged, thanks. Yirba (talk) 17:45, 26 July 2026 (UTC)Reply

Experimental notability evaluation tool

[edit]
Latest comment: 4 hours ago2 comments2 people in discussion

I’ve been working on a Toolforge tool that evaluates automatically detectable evidence relevant to notability and displays the results as a compact visual indicator. The tool is still experimental.

The tool is intended to help editors quickly see what types of evidence are present on an item. It is an aid to human reviewers, not a replacement for them. It cannot determine whether an item actually meets Wikidata's notability policy, and it can only evaluate information currently present in the item.

Documentation, installation instructions, known limitations, and a feedback page are available at User:Bovlb/wd-notability.

I’m particularly interested in reports of:

  • evidence that the tool failed to detect
  • evidence that the tool evaluated incorrectly
  • user interface or usability issues

This implementation is intended to reflect the current Wikidata notability policy, informed by community discussions and existing editorial practice. If you want to discuss Wikidata’s notability policy in general, that is probably better placed at the current RFC. Bovlb (talk) 19:21, 26 July 2026 (UTC)Reply

Thanks for building this. It took me a little while to figure out how it works, but fortunately the documentation is very helpful. It's a good way to identify items that could do with more work. --Yirba (talk) 23:12, 26 July 2026 (UTC)Reply

Remove protection at Q156723 (Louise of Mecklenburg-Strelitz)

[edit]
Latest comment: 6 hours ago2 comments2 people in discussion

Can you please remove protection of Louise of Mecklenburg-Strelitz ~2026-41429-58 (talk) 19:33, 26 July 2026 (UTC)Reply

Why? If you need to edit the item, you can request someone else to do it here. Samoasambia ✎ 20:56, 26 July 2026 (UTC)Reply
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat&oldid=2523212170"