close
Jump to content

Wikidata:Project chat

Edit links
Add topic
Shortcuts: WD:PC, WD:CHAT, WD:?
From Wikidata
Latest comment: 3 hours ago by Immanuelle in topic Support for RDF-star


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

Modern Abdera vs Ancient Abdera

[edit]
Latest comment: 2 days ago5 comments2 people in discussion

Hi!

I noticed the language links we're off on wikipedia:Abdera,_Thrace because some language links are tied to Q188615 (Abdera modern city) and Q132527193 ancient Abdera. The Wikipedia articles seem to cover both in the languages I have checked.

Should modern and ancient Abdera be represented by two different Wikidata items? Is there a guideline on this? Should it in that case be linked somehow? Which version should the pages that cover both be linked to?

I notice that a separation has been made between Q18287233 (ancient Rome) and Q220 (modern Rome) with a statement linking them, but encyclopedic importance of Rome could motivate special solutions for that city that might not be generalisable.

Our end goal should be to fix the language links on Wikipedia at least. (Note: there is also --Orubblig (talk) 21:47, 3 March 2026 (UTC)Reply

There is also an item for the municipality, so there are three items Ancient Abdera (Q132527193), Abdera (Q188615), and Avdira Municipality (Q4258195). German, Greek and Russian have articles for all three, so it is useful to have separate items. From the English article, it seems that the locations of ancient Abdera and modern Avdira are different. I haven't checked, but I would not be surprised if some articles were linked to the wrong item. TSventon (talk) 23:23, 3 March 2026 (UTC)Reply
That was a very good point. Tack! This demonstrates that there is need for separate items. On closer inspection I can now see that both are stated to be in Avdira Municipality and that they have been marked as different from each other. Is there a more suitable property? Rome vs Ancient Rome used said to be the same as. Which is correct?
Now that it is established that they should be separate articles, we have to take a decision on which is the most appropriate for each language article. For English, I would say it has to be moved to the modern city as the article is the modern city with ancient city as "history".--21:03, 4 March 2026 (UTC) Orubblig (talk) 21:03, 4 March 2026 (UTC)Reply
I am not an expert on when to use which property on Wikidata. The relations between ancient and modern cities seem to be different for Rome and Abdera, so different properties are probably appropriate. Rome has been in the same place and kept the same name since antiquity, while according to the en Wikipedia the ancient city of Abdera was abandoned, then a new settlement known as Polystylon emerged and was eventually given the name of Abdera. The en article says it is about the municipality which suggests a link to Avdira Municipality (Q4258195). TSventon (talk) 14:59, 5 March 2026 (UTC)Reply
It says it is about the municipality, but most of the article is about the ancient city. I wish there was a strict policy here to guide us. As you point out, it wouldn't really be accurate to point it to the modern city/town either as it is a renamed separate city.
Perhaps the solution is to create separate Wikipedia articles. In the meanwhile, I am leaning towards doing nothing. Sorry about dragging you in to this.--Orubblig (talk) 21:54, 11 March 2026 (UTC)Reply

Help: Service number for Royal Air Force personnel

[edit]
Latest comment: 1 day ago4 comments3 people in discussion

Dear Wikidata experts,

Wikidata currently recognises the property ‘United States Armed Forces service number (P2028)’, which is an instance of ‘service number (Q7455778)’. As I would also like to record this property for personnel of, for example, the Royal Air Force and Royal Australian Air Force, who have also been assigned service numbers, I wonder whether there have been any discussions in the past about this potential extension that resulted in a negative decision. Are there any reasons against extending the property ‘service number (Q7455778)’ to other military units?

Best regards,

Rico Ricodiparma (talk) 18:43, 5 March 2026 (UTC)Reply

Presumably they would not be globally unique identifiers, which is a requirement for external id's. Just propose new properties for the other service numbers, identifiers like that are pretty routine here. ArthurPSmith (talk) 14:38, 6 March 2026 (UTC)Reply
There was Wikidata:Property proposal/Service number, which was only opposed because of a suggestion to repurpose P2028 by changing its labels, descriptions, constraints, URLs and data type and requiring qualiifiers; it would have been easier to create a new property. Peter James (talk) 19:53, 6 March 2026 (UTC)Reply
Ok, thank you both for your suggestions! I will try to propose my first property. Ricodiparma (talk) 14:34, 12 March 2026 (UTC)Reply

Creating 2 different Wikipedia links, each in a separate Wikidata item, which redirect to the same page

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

wikipedia:Time average redirects to wikipedia:Moving average, but actually they are 2 related albeit different concepts. How should I go about adding the Wikipedia link to my newly-created time average - Wikidata item? I'm sure there's a way... AddyLockPool (talk) 10:20, 7 March 2026 (UTC)Reply

Wikidata:Sitelinks to redirects#Using badges to enter a sitelink to a redirect. Basically add the link using Wikidata UI as usual, but also include the intentional sitelink to redirect (Q70894304) badge. --Matěj Suchánek (talk) 14:31, 7 March 2026 (UTC)Reply

Adding new lexemes by bot

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

We're preparing a batch of Moroccan Arabic lexemes (as part of this research project), but it seems that the Quickstatement tool does not allow adding new lexemes. Can I directly use my bot DarijaBot to perform this new task, or do I have to make a new bot request? Ideophagous (talk) 19:41, 7 March 2026 (UTC)Reply

It looks like the Wikidata_Lexeme_Forms tool have a bulk mode that might be used. I've never tried it, so I don't know if it fits your use-case. I would advise to perform the edits using the bot account just in case there are any issues with your batches which would need reverting or even blocking. All self-made bots needs approval. If you're familiar with Javascript I'm sure Wikibase-CLI would appreciate getting a patch to implement support for creating lexemes, this way other people would benefit. Certain general bulk-editing tools that implement EditGroups can be used without getting approval. Infrastruktur (talk) 14:39, 8 March 2026 (UTC)Reply
Hello. Thanks for the information. The bot already operates on Wikidata. I just want to know if I can extend its operations to this new task, or if each new task should be approved. I will check the tool you suggested. Cheers! Ideophagous (talk) 20:40, 8 March 2026 (UTC)Reply
Each new task should be approved, yes. ArthurPSmith (talk) 16:36, 9 March 2026 (UTC)Reply

Script to make bulk RfDs

[edit]
Latest comment: 4 days ago3 comments2 people in discussion

Do we have such a script? MediaWiki:Gadget-RequestDeletion.js only allows single requests. NguoiDungKhongDinhDanh 11:58, 8 March 2026 (UTC)Reply

There currently isn't a script. People usually go to https://www.wikidata.org/w/index.php?title=Wikidata:Requests_for_deletions&action=edit&section=new&preload=Wikidata:Requests%20for%20deletions/Preload_2&editintro=Wikidata:Requests_for_deletions/Editintro_2&nosummary=1 and fill in the blanks to create a bulk request. Ternera (talk) 13:11, 9 March 2026 (UTC)Reply
I'm aware of that form. It isn't very convenient, though. NguoiDungKhongDinhDanh 16:52, 9 March 2026 (UTC)Reply

Wikidata weekly summary 722

[edit]
Latest comment: 4 days ago1 comment1 person in discussion
Image
Here's your quick overview of what has been happening around Wikidata in the
week leading up to 2026-03-09. Missed the previous one? See issue #721.

Discussions

  • New requests for permissions/Bot: DifoolBot 8 - Task: Remove Wikipedia import references from statements where the referenced article has since been deleted.

Events

Press, articles, blog posts, videos

Tool of the week

  • Data2Go : An iOS app for browsing and editing Wikidata with a mobile-first UI. It combines fast search, rich item detail views, statement editing, map previews, media galleries, and account-based write access to Wikidata.

Other Noteworthy Stuff

Newest properties and property proposals to review

You can comment on all open property proposals!

Did you know?

Development

  • GraphQL: We have released the first version of this new API. You can read more about it and try it out.
  • Wikidata integration in the other Wikimedia projects: We have worked with editors to improve some highly-used modules that access Wikidata in a suboptimal way. They have been adjusted to lead to less unneeded change notification in people's watchlists and recent changes on Wikipedia and co.
  • Mobile statement editing: We fixed an issue where certain Properties were not accessible in the new mobile UI. We are also working on fixing a bug with saving musical notation statements. We are continuing the work on supporting editing of geocoordinates.

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

Latest population numbers

[edit]
Latest comment: 4 days ago1 comment1 person in discussion

@AniaGrzybowska, Unpocoloco: thanks for importing population numbers to items. Could you please change the rank of the latest value to preferred so that this number is displayed eg in dynamic tables etc? Prior discussion was at Wikidata:Request a query/Archive/2026/02#Only latest population number (ping@Piastu:). It may be good if some script / bot auto-changed the rank of latest population numbers to preferred. Prototyperspective (talk) 18:51, 9 March 2026 (UTC)Reply

Temporary Accounts replacing IP editing for privacy reasons

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

Why would Temporary Accounts replace IP editing for privacy reasons? ~2026-15304-43 (talk) 18:07, 10 March 2026 (UTC)Reply

They already have. Ainali (talk) 19:33, 10 March 2026 (UTC)Reply
See Help:Temporary accounts. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:15, 11 March 2026 (UTC)Reply

Making a property more broad

[edit]
Latest comment: 3 hours ago2 comments1 person in discussion

I realize that a property that was made several months ago (‎modern shrine ranking (P13723)) can be made more broad, and with absolutely zero ground being made on another related property, I would like to broaden the existing category to all shrine ranks, and then use a qualifier to indicate which shrine ranking system is being represented. And adding a mandatory conferred by (P1027) qualifier to indicate the ranking system in question.

Rename ‎modern shrine ranking (P13723) to "shrine ranking" and every existing usage of the property will have conferred by (P1027) modern system of ranked Shinto shrines (Q712534). For things covered by the other proposal they will get conferred by (P1027) Engishiki ranking (Q138640329).

Is this something we can do? Immanuelle (talk) 21:23, 10 March 2026 (UTC)Reply

Still wanting to know if this can happen Immanuelle (talk) 03:45, 14 March 2026 (UTC)Reply

We used Wikidata to build a 33M image dataset for training vision models (EntityNet)

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

Hi all,

I'm one of the authors of a recent paper in which we made heavy use of Wikidata to create a large-scale image-text dataset for training CLIP vision-language models. I thought this community might find the approach interesting, since it is built directly on Wikidata’s structure.

What we did: We used SPARQL queries over the subclass-of (P279) and parent-taxon (P171) hierarchies to extract approximately 135k visual entities across 21 super-entity categories, including animals, plants, tools, and buildings. For each entity, we collected its name, description, aliases, and sitelink count. We also extracted attributes from Wikidata related to color, partonomy, behavior, and other aspects, and generated and curated additional attributes using LLMs. These entities and attributes were then used to generate search queries for downloading images from the web.

The resulting EntityNet dataset contains 33M images and 45M alt texts. The images are linked to 135k Wikidata entities, and in total we exported 613k text labels from Wikidata. One key finding was that training on a mix of Wikidata’s structured entity information and noisy web alt texts works better than using either alone: the knowledge graph metadata improves model quality.

The dataset, including image URLs and curated tables of entities and attributes, along with the paper, code, and trained models, is openly available here: https://entity-net.github.io/

Check it out if it sounds interesting. I’m happy to answer any questions. Simon-ging (talk) 11:39, 11 March 2026 (UTC)Reply

WDQS Triple Store Evaluation Benchmark Results Report from Wikidata Platform Team

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

As mentioned in the16 February 2026 Wikidata status update, the Wikidata Platform Team has completed an initial evaluation of alternatives to Blazegraph for the Wikidata Query Service (WDQS) and published a report. Anyone who uses the WDQS should read this report and decide whether they believe that the conclusions drawn there are supported by the evaluation that the team performed.

Because the report is a PDF document there is no obvious place to make comments. Some comments have been placed in the Discussion area of the team's monthly newsletter. (Be warned that this link is for the current-at-this-time newsletter and based on past practices the February newsletter may move to February newsletter.) Peter F. Patel-Schneider (talk) 12:59, 11 March 2026 (UTC)Reply

Question about item qualifier

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

Say I have an item with instance of (P31) religious organization (Q1530022). What qualifier property that would be appropriate to denote its religion or maybe ethnicities? also is there any item example that does have the appropriate qualifier? FenyMufyd (talk) 02:08, 12 March 2026 (UTC)Reply

religion or worldview (P140) is the one I use. I don't know of any very good examples but Society of Jesus (Q36380) might be helpful. Secretlondon (talk) 13:08, 12 March 2026 (UTC)Reply

Merge request for Rafa al-Yadayn

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

Hello, could someone please merge Raf al-Yadayn (Q131418164) (English/Bengali) and Raf al-Yadayn (Q61970470) (Urdu/Punjabi)? They are duplicates of the exact same topic. Thank you. Ch.AhmedRaza23 (talk) 07:58, 12 March 2026 (UTC)Reply

Someone has done this for you. Secretlondon (talk) 13:12, 12 March 2026 (UTC)Reply

Adding data for Attest - consumer research engine

[edit]
Latest comment: 19 hours ago3 comments2 people in discussion

Hi there,

I am looking to add reliable unbiased factual data about Attest: https://www.askattest.com/

It is a consumer research engine for B2C brands, with customers including Netflix, Reddit, Unilever, Burger King.

For full transparency I am an independent comms consultant working with Attest on a short term basis.

Many thanks Abigail Insight (talk) 10:24, 12 March 2026 (UTC)Reply

See WD:SELF and WD:N as to why we aren't keen on self promotion of commercial organisations. Vicarage (talk) 11:06, 12 March 2026 (UTC)Reply
Understood. For background, the company has been around for ten years and to date, Attest has raised >$100 million in venture capital with backing from investors such as Kismet, Schroders, and NEA.
Here are some notable articles: https://techcrunch.com/2021/10/27/attest-raises-60m-to-expand-its-no-code-approach-to-research-surveys-as-a-service-which-taps-110m-consumers-in-49-countries/
https://www.cityam.com/london-tech-startup-attest-closes-16m-groupon-investor-nea/
https://www.research-live.com/article/news/attest-develops-ai-tool-for-researchers/id/5146183 Abigail Insight (talk) 12:38, 13 March 2026 (UTC)Reply

truncated image names with filetype changed

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

HMS Grafton (Q5632604) has an image (P18) "HMS Grafton, fitted with a jury rudder, etc, for her voyage to England, after the storm off Louisbourg, 1757.".png, which is surely wrong, as the actual image on Commons is https://commons.wikimedia.org/wiki/File:HMS_Grafton,_fitted_with_a_jury_rudder,_etc,_for_her_voyage_to_England,_after_the_storm_off_Louisbourg_1757200500089.jpg which has the filetype changed from jpg to png, and the filename truncated. When I partiallly delete, and let it autocomplete, I get the png version. It displays the image in the page, so it still knows about it. What's going on? Vicarage (talk) 16:41, 12 March 2026 (UTC)Reply

Both images exist, from different sources. Remove the ".png" and both versions are there, the .png first and .jpg second. The .png is from a book and has the caption below. Peter James (talk) 20:22, 12 March 2026 (UTC)Reply

Weird diffs appearing in Wikipedia

[edit]
Latest comment: 18 hours ago1 comment1 person in discussion

Hello Wikidata,

In my WP watchlist, I have been enabling the "see Wikidata modifications" filters, and I (and another user) have noticed some modifications from WD elements appearing in my watchlist. The details are as follows : I follow fr:Márta Svéd, linked to Q51565983, not her husband George (Q135523780), but I saw this diff in my watchlist. Even weirder, I follow fr:Jacques-Germain Soufflot (Q550244) and received this diff of Q56352703 (which I don't follow, it is not linked to any fr:wp article). Even weirder, I received only this diff, none of the other modifications on this element the same day, nor the previous one from the 25th Feb. Any idea on the reason ? Cuagga (talk) 12:50, 13 March 2026 (UTC)Reply

Support for RDF-star

[edit]
Latest comment: 3 hours ago1 comment1 person in discussion

I think we should be moving towards supporting RDF-star. At least making it so that qualifiers can be queried as triple attached to a triple. Is there any work being done on supporting this? Immanuelle (talk) 03:48, 14 March 2026 (UTC)Reply

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat&oldid=2471777656"