Shortcuts: WD:PC, WD:CHAT, WD:?

Wikidata:Project chat

From Wikidata
(Redirected from Wikidata:Help desk)
Jump to navigation Jump to search

Wikidata project chat
A place to discuss any and all aspects of Wikidata: the project itself, policy and proposals, individual data items, technical issues, etc.

Please use {{Q}} or {{P}} the first time you mention an item or property, respectively.
Other places to find help

For realtime chat rooms about Wikidata, see Wikidata:IRC.
Start a new discussion
Edit
On this page, old discussions are archived after 7 days. An overview of all archives can be found at this page's archive index. The current archive is located at 2024/05.

Long-term vandalism 62.43.191.154[edit]

Found at https://www.wikidata.org/w/index.php?title=Q30541537&diff=1905583343&oldid=1901236773

Each day edits are made from this IP there seems to be vandalism involved https://www.wikidata.org/wiki/Special:Contributions/62.43.191.154 Rostworowski (talk) 19:17, 11 May 2024 (UTC)[reply]

@Rostworowski his last edit was on 24 April 2024. Hence - stale request. Next time please post vandalism-related messages to Wikidata:Administrators' noticeboard Estopedist1 (talk) 07:08, 13 May 2024 (UTC)[reply]
"his last edit was on 24 April 2024. Hence - stale request." - waht he, what request, what stale? Your message doesn't make any sense at all. Rostworowski (talk) 13:39, 14 May 2024 (UTC)[reply]
It does. Sanctions against IP addresses are seldom done in retrospection. They are meant to be protection, not punishment. And yes, your request should go to WD:AN. --Matěj Suchánek (talk) 06:24, 16 May 2024 (UTC)[reply]
No it doesn't. What sanctions do you talk about? And why wouldn't it be "protection" but "punishment" if this IP would be blocked? "And yes, your request should go to WD:AN" - which request? Not even one of your four sentences makes any sense. Rostworowski (talk) 23:39, 18 May 2024 (UTC)[reply]

Network or Service Problems[edit]

When the service is fluctuating or absence for some days as in many towns in Nigeria how could someone edit or create a page? Rabiu tijjani (talk) 20:55, 11 May 2024 (UTC)[reply]

Is overthrowing the government of Nigeria cheaper than getting Starlink? Infrastruktur (talk) 20:48, 13 May 2024 (UTC)[reply]
Here in Nigeria, we are not overthrowing our government. Nigeria is a republican country, we are changing the government through election for every four (4) years period. Starlink may be get this or next government. Maibadali (talk) 15:57, 19 May 2024 (UTC)[reply]
It was said in jest, but the background for it is this: https://techpoint.africa/2019/11/28/nigerias-social-media-bill/ . Infrastruktur (talk) 19:04, 19 May 2024 (UTC)[reply]
good Maibadali (talk) 19:17, 19 May 2024 (UTC)[reply]

Search priority[edit]

Hello,

for a while I have been wrapping my head around how Wikidata prioritizes entities when performing text search on the main Wikidata page (the input box in the right upper corner, e.g. why when writing human the results show human (Q5) on the first line and not something different). I have made a comment few weeks back, and was made familiar that Wikidata uses Elastic Search and probably this is the code mirror repository. I have been looking into it and found profile config files. But I still don't quite understand.

There are profiles for the EntitySearch and EntityPrefixSearch, which to my understand should be the main queries when performing search. Then there are the rescore config files, to my understanding rescore in Elastic is performed on results yilded by the main query. Looking into the files, the rescore functions prioritize entities that have many incoming links and sitelinks (two saturation functions with weights 0.6 and 0.4 while weight of query is 1 and weight of rescore query is 1, the final score should be sum of those scores), but I could not find where the rescore functions are used and whether there are more used somewhere (e.g. appending to a chain of rescore functions). Furthermore, the Elastic bm25 scores are not normalized to [0,1], is there some normalization step before rescoring? Since the rescore with incoming and sitelinks is in the interval [0,1].

Is there someone who would be please willing to explain the process step by step? Martin Gora (talk) 11:28, 13 May 2024 (UTC)[reply]

@Martin Gora: I don't think the developers are regularly checking this board. Maybe try Wikidata:Report a technical problem? ArthurPSmith (talk) 17:23, 13 May 2024 (UTC)[reply]
But “I’m curious how this works” is hardly a technical problem IMHO. (Also, given that it’s search-related, Wikidata:Report a technical problem/WDQS and Search would be a slightly better fit than the parent page.)
@Martin Gora: The search team holds monthly “Talk to the Search Platform / Query Service Team” open meetings (usually on the first Wednesday of each month; see May announcement), I think those would be a good venue for your question. Lucas Werkmeister (WMDE) (talk) 09:03, 14 May 2024 (UTC)[reply]
@Martin Gora CirrusSearch makes use of search profile to tune the retrieval and ranking functions depending on the use-cases. The list of search profiles can be extracted using the cirrus-profiles-dump API.
The search profile system has different types (similariry, ft_query_builder, rescore to name a few), knowing how they get assembled is far from trivial but you could get a sense by looking at the contexts section of the API results above, for instance wikibase_fulltext_search is the search context we use to rank results in Special:Search when selecting namespaces holding wikidata entities, if searching for other namespaces we fallback to the default context.
If you are curious you can see extract some debug information from the search system:
  • see how the query is actually built by appending &cirrusDumpQuery to your search URL
  • see how the explanation of how the scores are computed by appending &cirrusDumpResult&cirrusExplain=pretty to your search URL
  • you can also instruct CirrusSearch to a use particular profile by appending the uriParam listed in the overriders from the profiles API, e.g. you can force to use another rescore function by appending &cirrusRescoreProfile=profile_name to your search URL
We generally do not normalize the retrieval scores before applying the rescore window functions indeed, given the wide variety of fields we use, it would be hard, I think, to get to reasonable normalization function. This means that the more words you have in your query the more impactful the text matches will be, which I think is reasonable. For wikis where we have enough data we train a model that results in a decision tree (Q831366) which is less prone to normalization problems.
As suggested by Lucas please feel free to join one of our office hours we would be happy to explain this to you in more details. DCausse (WMF) (talk) 19:39, 14 May 2024 (UTC)[reply]
Hello,
I want to thank you - @ArthurPSmith, @Lucas Werkmeister (WMDE), and @DCausse (WMF)) - for your helpful comments. I will definitely try to join the montly meeting.
Best regards, Martin Gora (talk) 11:56, 18 May 2024 (UTC)[reply]

Zhakiyanov (Q42309946)[edit]

What should be done with this incorrect item ? Not a given name but a surname as the English Wikipedia article says. Kpjas (talk) 19:06, 14 May 2024 (UTC)[reply]

✓ Done LockaPicker (talk) 16:42, 18 May 2024 (UTC)[reply]

Coordinates display in decimal[edit]

I prefer to add coordinates in decimal format and would like to see them in decimal rather than DMS format. Is there a gadget like thing that can help with this? Arjunaraoc (talk) 14:56, 15 May 2024 (UTC)[reply]

I'm not aware of one. If you want to write your own, I'd start by looking at Wikidata:Tools/Enhance_user_interface#Userscripts_to_enhance_the_display_of_values. Bovlb (talk) 15:19, 15 May 2024 (UTC)[reply]
Why is the default DMS? There is en:ISO 6709. If I understand it right, they say: "Fraction of degrees (decimal degrees) is preferred in digital data exchange, while sexagesimal notation is tolerated for compatibility". Please can we change the default? --sk (talk) 09:29, 16 May 2024 (UTC)[reply]
Sexagesimal is easily divisible for humans. Coordinate notation in DMS is mostly historical at this point, made irrelevant by the advent of computers. And I still see coordinates presuming to be accurate to 10000th of a second on Wikidata, which is dubious seing as commercial GPS is accurate to between 1-10 meters. If mathematicians ruled the world we'd probably have coordinates based on radians instead. :P Infrastruktur (talk) 10:14, 16 May 2024 (UTC)[reply]
I'd vote to change the default to decimal. Its easier to mentally compare 2 numbers at the precision level I'm interested in, and copy the information out to other places. Vicarage (talk) 10:36, 16 May 2024 (UTC)[reply]
I support. Arjunaraoc (talk) 14:55, 16 May 2024 (UTC)[reply]
I would have no objection to switching to decimal. Bovlb (talk) 03:30, 17 May 2024 (UTC)[reply]
"digital data exchange" means things like APIs and file formats (things designed for computers to read). How it should be displayed for humans to read is a different matter, which that page covers in the section "Representation at the human interface (Annex D)". - Nikki (talk) 15:13, 20 May 2024 (UTC)[reply]
@Nikki, Thanks for providing the link, which helps clarify the current situation. I think a gadget is certainly required to add adding decimal display a, as it is more convenient and less error prone for humans also to work with. Arjunaraoc (talk) 00:35, 23 May 2024 (UTC)[reply]
Thanks to @RamSeraph for implementing the request as part of his gadget. Arjunaraoc (talk) 08:34, 25 May 2024 (UTC)[reply]

Geolocate[edit]

Hi, I want to create a new statement about "Battle of Seseña" (https://www.wikidata.org/wiki/Q923995), because i want to include a "coordinate location". I don´t know how to do it, that's the reason i'm writting this. Thank you in advance. Elena ElenaGarciadeDiego (talk) 10:39, 17 May 2024 (UTC)[reply]

On mobile it is not implemented to add statements yet, on desktop see Help:Statements#Adding_statements LockaPicker (talk) 12:17, 17 May 2024 (UTC)[reply]
You don't need to create a new object for the location, just enter them in the format 52°30'59"N, 13°22'40"E LockaPicker (talk) 11:47, 18 May 2024 (UTC)[reply]
Oh okay, it was easier than i thought. Thanks!! ElenaGarciadeDiego (talk) 07:33, 20 May 2024 (UTC)[reply]

Query Service and unknown for dates of birth and death[edit]

I did some albeit non-exhaustive searching before posting this.

If a value for "unknown" is added to the "date of birth" or "date of death" fields for a person's wikidata item, the Query Service presents in the query output an odd url with a base of http://www.wikidata.org/.well-known/genid/ and a terminal value which looks like a checksum, for example "792cc70929f20df5f1726b74e278a715". The composed URLS always return a "page not found" error. The query I pulled this particular ID from is shown below in a pre/pre block. What I'm wondering is whether this expected behavior or not. Maybe it is a known bug that has not been addressed? Thanks for considering this question. --Ceyockey (talk) 02:13, 18 May 2024 (UTC)[reply] 

SELECT ?person ?article ?DOB ?DOD ?DESC
WHERE
{
  ?person wdt:P31 wd:Q5 .   # human
  ?person wdt:P734 wd:Q5195456 .
  ?article schema:about ?person . 
  ?article schema:isPartOf <https://en.wikipedia.org/>.
  OPTIONAL {?person wdt:P569 ?DOB }
  OPTIONAL {?person wdt:P570 ?DOD }
  OPTIONAL {?person schema:description ?DESC}
  FILTER(LANG(?DESC) = "en")
}
ORDER BY ?article

Ceyockey (talk) 02:13, 18 May 2024 (UTC)[reply]

https://www.wikidata.org/wiki/Wikidata:Request_a_query has helpful people Vicarage (talk) 11:53, 18 May 2024 (UTC)[reply]
I think this is just an oddity of the way we store unknown values. See mw:Wikibase/Indexing/RDF_Dump_Format#Special_values. Also, you might like to use the {{SPARQL}} template when quoting SPARQL. Bovlb (talk) 16:58, 18 May 2024 (UTC)[reply]

Opinions on data structure[edit]

So, I wanted to add the position in the world ranking list for top 100 chess players each month. Now I wonder what structure should be preferred.

Magnus Carlsen (Q106807)Elo rating (P1087)2906point in time (P585)+2024-04
Magnus Carlsen (Q106807)ranking (P1352)1determination method (P459)Elo rating system (Q105955)point in time (P585)+2024-04

or

Magnus Carlsen (Q106807)Elo rating (P1087)2906ranking (P1352)1point in time (P585)+2024-04

LockaPicker (talk) 11:24, 18 May 2024 (UTC)[reply]

I know little about chess, but the first seems sufficient. The elo ranking can be deduced from a sorted query, and it would not make rankings based on other ratings to be harder to query Vicarage (talk) 11:56, 18 May 2024 (UTC)[reply]

After the mayor kitty …[edit]

the graduate cat ! @Denny:. author  TomT0m / talk page 17:49, 18 May 2024 (UTC)[reply]

That's great news! No Wikidata item for Dr Dow yet? :D --Denny (talk) 18:07, 20 May 2024 (UTC)[reply]

Priority in alias handling[edit]

Using this as an example, and the related english article: if you have two names that refer to the same concept (oil body and oleosome), with only one present as a page in en:wikipedia (oil body), is it sufficient to add the synonym (oleosome) as an alias to the corresponding wikidata item (by doing this I was able to search for "oleosome" using the internal search function and have the suggestion for "oil body" pop up), or should I also create a redirect page (oleosome -> oil body)? Salusbibbi (talk) 18:34, 18 May 2024 (UTC)[reply]

edit: I noticed the redirect I was talking about already exists, and this was probably the reason for the search results. Anyway, the question still stands in general, but I imagine the indexing in the internal search (and possibly also external) comes exclusively/preferentially from redirects, instead of aliases, so the former should be preferred. Am I right? Salusbibbi (talk) 18:57, 18 May 2024 (UTC)[reply]
Search in Wikidata is independent from Wikipedia pages. Aliases matter for your Wikidata search but don't have an effect on what you see when you search within Wikipedia. ChristianKl ❪✉❫ 15:44, 19 May 2024 (UTC)[reply]

"author" and "author name string"[edit]

If "author" is added, is it necessary to remove "author name string" ? Or will a bot do it ? Io Herodotus (talk) 19:29, 18 May 2024 (UTC)[reply]

Does it need removing at all? It seems to me that capturing how the author is identified in the referenced work itself may be important, even if the author has an item. M2Ys4U (talk) 20:31, 18 May 2024 (UTC)[reply]
@M2Ys4U, yes, it should be removed. If you're able to include author (P50), then add subproperty object named as (P1932) with the name as given in the source. — Huntster (t @ c) 17:08, 19 May 2024 (UTC)[reply]
Generally, I would only expect only one of either for a series ordinal (P1545) value, but to have both isn't a thing that needs to be fixed. Also the author (P50) should have all the same references and qualifiers from the original. In addition, the new statement should have a qualifier of object named as (P1932) with the value of whatever was the previous value of author name string (P2093).
Here is an example of how that might look: [1]. In that example I'm using the tool Author Disambiguator (Q76693569) to make the update. It's a nice tool that can combine both instances.
In regards to whether there are bots that automatically do that, I don't think there are because often times scientific articles have data imported from two sources where the author data is different, mostly when different editions of the same work are erroneously merged. William Graham (talk) 21:05, 18 May 2024 (UTC)[reply]
I thought "author" means : there is an identified page
"author name string" means : there is no page yet Io Herodotus (talk) 21:11, 18 May 2024 (UTC)[reply]

Merge "Galician films" and "Filmes de Galicia" categories[edit]

Hello everyone, I'd like to request the merge of the following items on Wikidata: Q125962414 (Category:Galician films) Q9908259 (Categoría:Filmes de Galicia); Both refer to the same category and should be merged to avoid duplications. Additionally, I'm encountering issues when trying to add the tag "Category:Filmes de Galicia" to item Q125962414 due to the pre-existence of this tag in Q9908259. I would greatly appreciate it if an administrator could assist in merging them. Thank you for your help! 92.176.79.149 21:12, 18 May 2024 (UTC)[reply]

✓ Done (no need to be admin). VIGNERON (talk) 08:55, 22 May 2024 (UTC)[reply]

Merging Category:Food of the United States into Category:Food and drink in the United States[edit]

Category:Food of the United States (Q111318637) should be merged into Category:Food and drink in the United States (Q8463135), but the merge failed when I tried to do that. jlwoodwa (talk) 04:08, 19 May 2024 (UTC)[reply]

Same for Category:Kirksville, Missouri (Q65735906) into Kirksville (Q966460). jlwoodwa (talk) 05:51, 19 May 2024 (UTC)[reply]
I'm not sure if the first should be merged as one is "food" and the other is "food and drink". The second should not be merged as one is a category and the other is not. If the only sitelink was Commons it could be moved to Q966460 and then Q65735906 could be deleted, but there is also a German Wikipedia category that would not be suitable for Q966460 (and could not be merged as it conflicts with the sitelink already there). Peter James (talk) 17:03, 19 May 2024 (UTC)[reply]
Agree, both should definitely not be merged. Sjoerd de Bruin (talk) 07:36, 22 May 2024 (UTC)[reply]

Can grants fully fund a scholarly article and are there any examples in Wikidata?[edit]

I'm interested in if a grant can sort of fully fund a scholarly article so that I could add funder (P8324)...or at least correlate a funding grant to a scholarly article. I looked at an orcid of a researcher and under funding I found many grants...but I had no idea if these can correlate to any actual scholarly papers. It looks more like grants are something like "we plan to do some things" but I get the impression that can't correlate to any research...

Is there any data on Wikidata where specifical scholarly articles were funded by grants and can you please link me a few scholarly articles here so I can see them.

Maybe I need to do some work, I don't know interview some charities and researchers who provide grants to get to this information?

Is this data simply not available? If yes, is there a reason why? AutismresearchNeedsupdates (talk) 08:23, 19 May 2024 (UTC)[reply]

Let's look at the example where the NIAID funded Coronavirus gain-of-function research in Wuhan. Baric, Shi et al published A SARS-like cluster of circulating bat coronaviruses shows potential for human emergence (Q36702376). In it, they have a Acknowledgements section in which they say: "Research in this manuscript was supported by grants from the National Institute of Allergy & Infectious Disease and the National Institute of Aging of the US National Institutes of Health (NIH) under awards U19AI109761 (R.S.B.), U19AI107810 (R.S.B.), AI085524 (W.A.M.), F32AI102561 (V.D.M.) and K99AG049092 (V.D.M.), and by the National Natural Science Foundation of China awards 81290341 (Z.-L.S.) and 31470260 (X.-Y.G.), and by USAID-EPT-PREDICT funding from EcoHealth Alliance (Z.-L.S.)."
If you Google for U19AI109761 you find https://govtribe.com/award/federal-grant-award/cooperative-agreement-u19ai109761 where the grant gets explained but we currently don't have a Wikidata item for that grant. It would be possible for someone to upload those federal grants to Wikidata and run an LLM on the Acknowledgements sections of papers to get the data to link it to papers. ChristianKl ❪✉❫ 09:02, 19 May 2024 (UTC)[reply]

Protect Q1922726[edit]

My edit of this page get regularly reverted by someone without an account. they remove the abolishment date even though I have sources. Could it be possible to protect this page? Jhowie Nitnek 08:49, 19 May 2024 (UTC)[reply]

request[edit]

hi, please can someone link en:twitch streamer to Twitch streamer (Q50279140) with intentional sitelink to redirect (Q70894304)? i'm an IP address so it's not letting me add it with the badge. 82.132.221.123 13:45, 19 May 2024 (UTC)[reply]

Done. ChristianKl ❪✉❫ 15:11, 19 May 2024 (UTC)[reply]

Logainm ID[edit]

I merged Ballyforan (Q4076806) and Ballyforan (Q104337829) but now I realize they had different Logainm IDs with different outlines, can someone figure out what distinguishes the two entities so I can restore Q104337829. RAN (talk) 18:20, 19 May 2024 (UTC)[reply]

If you look at the maps at Logainm it seems like one is bigger than the other. Besides that ChatGPT has to say about the difference "The second link, https://www.logainm.ie/ga/42580, provides more detailed historical context regarding the place name "Béal Átha Feorainne." It includes various historical forms and references, such as different name spellings from records dating back to the 16th and 17th centuries, and linguistic insights into the name's evolution. This additional information is not present in the first link, which focuses more on the modern aspects and current usage of the place name." ChristianKl ❪✉❫ 21:21, 19 May 2024 (UTC)[reply]
https://www.logainm.ie/ga/42580 is the townland (townland (Q2151232)), and https://www.logainm.ie/en/1414615/ is the population centre (a village, most of which is in the townland, with a few buildings in Coolatober (Q104337977)). I don't know whether to combine these and use subject has role (P2868) on the identifiers or keep them separate - civil parishes and villages in England are usually combined even if a few houses in the village are outside the parish boundary, but in Ireland there can be a townland, civil parish (containing several townlands) and village all with the same name, and Logainm has identifiers for each but sometimes an identifier does not distinguish. Peter James (talk) 15:34, 20 May 2024 (UTC)[reply]

What to do with Karim Ahmad Khan (Q105209571)?[edit]

I was updating Karim Ahmad Khan (Q6370541) and noticed there was an identical item at Karim Ahmad Khan (Q105209571). I was going to merge them but noticed that Karim Ahmad Khan (Q105209571) seems to originally have been another person which was merged into Tommy Jones-Davies (Q7819577). What's the recommended course of action? Piecesofuk (talk) 18:26, 20 May 2024 (UTC)[reply]

It was about another person for less than a minute after creation, and then a redirect for a few minutes before being used again (all by the editor who created the item). I don't think that's the same a repurposing an item that had existed for some time or was created by another editor - when duplicating an item it can take more than a minute to change the statements that need changing. Peter James (talk) 19:29, 20 May 2024 (UTC)[reply]
It's okay, somebody else has merged it now. I was under the impression that merged items should never be reused in any circumstances https://www.wikidata.org/wiki/Help:Merge (I noticed in this case that in Karim Ahmad Khan (Q105209571) most of the labels still referred to the original item https://www.wikidata.org/w/index.php?title=Q105209571&oldid=2140576186) Piecesofuk (talk) 08:15, 21 May 2024 (UTC)[reply]

Wikidata to split as sheer volume of information overloads infrastructure[edit]

From the 16 May 2024 issue of The Signpost (Q16639816)

Wikidata to split as sheer volume of information overloads infrastructure (Q126011233) Bluerasberry (talk) 14:00, 21 May 2024 (UTC)[reply]

  • It looks like scholarly articles will have their own database called Wikicite, the authors and topics will be linked to the Wikidata entries. --RAN (talk) 00:52, 23 May 2024 (UTC)[reply]
isn't this headline misleading? wikidata is not splitting. the query service is. BrokenSegue (talk) 01:23, 23 May 2024 (UTC)[reply]
I agree that the headline is misleading. The change is being made to the Wikidata Query Service, not Wikidata proper. It's unclear whether the author of the op-ed understands the nuance or not. -- William Graham (talk) 02:15, 23 May 2024 (UTC)[reply]

Wikidata weekly summary #628[edit]

Here's your quick overview of what has been happening around Wikidata over the last week.
This is the Wikidata summary of the week before 2024-05-20. Please help Translate.

Discussions

  • New requests for permissions/Bot:
    • Browse9ja Bot - Task: Automated data retrieval focusing on Nigerian / African-based information, integrating a chatbot, NLP API, knowledge graph, and machine learning model.
    • DhlizBot - Task: add Dhliz ID (P12696) for persons/films/TV series that exist in Dhliz (Q125594802) database.
    • AmeisenBot 2 - Task: Working on bot requests at WD:RBOT using QuickStatements.

Press, articles, blog posts, videos

Tool of the week

  • Wikidata GenderStats measures gender bias at the user level by computing stats on created items.
  • Created items : a list of items created by a user with labels. This not new but this has been repaired.

Other Noteworthy Stuff

Newest properties and property proposals to review

You can comment on all open property proposals!

Did you know?

Development

  • We have completed the preparation work to migrate the Query Builder's front end from Wikit to Codex (phab:T360096)
  • We also continued work on the EntitySchema data type and refining the CI flow for EntitySchemas (phab:T362000)

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

Feedback invited on Procedure for Sibling Project Lifecycle[edit]

You can find this message translated into additional languages on Meta-wiki. Please help translate to your language

Dear community members,

The Community Affairs Committee (CAC) of the Wikimedia Foundation Board of Trustees invites you to give feedback on a draft Procedure for Sibling Project Lifecycle. This draft Procedure outlines proposed steps and requirements for opening and closing Wikimedia Sibling Projects, and aims to ensure any newly approved projects are set up for success. This is separate from the procedures for opening or closing language versions of projects, which is handled by the Language Committee or closing projects policy.

You can find the details on this page, as well as the ways to give your feedback from today until the end of the day on June 23, 2024, anywhere on Earth.

You can also share information about this with the interested project communities you work with or support, and you can also help us translate the procedure into more languages, so people can join the discussions in their own language.

On behalf of the CAC,

RamzyM (WMF) 02:24, 22 May 2024 (UTC)[reply]

Attribution text[edit]

At Halsted House (Q96358236) "attribution text" is giving an error. When I looked to see how it was used at another The Musicians (Q655705) it is used the same way with the same error message, any ideas? RAN (talk) 05:12, 22 May 2024 (UTC)[reply]

The property is designed for use on structured data on Commons and should not be used on Wikidata itself. Sjoerd de Bruin (talk) 07:34, 22 May 2024 (UTC)[reply]
+1, this property was not meant for that (as indicated on the constraint itself). I see that DaxServer did a big import using this property wrongly. I guess it should be removed (or moved, maybe as a qualifier of collection (P195) ?). Also ping Fuzheado. Cheers, VIGNERON (talk) 08:52, 22 May 2024 (UTC)[reply]
Ah, I didn't notice the constraint. I'll remove it. But where should the credit line go, as described in MET object https://www.metmuseum.org/art/collection/search/132 ? DaxServer (talk) 09:23, 22 May 2024 (UTC)[reply]
collection doesn't seem to have the attribution text, but has a donated by (P1028). However, not all cases in the MET collection were donated with some being purchased. MET doesn't have that data categorized and just puts everything under "creditLine" DaxServer (talk) 09:27, 22 May 2024 (UTC)[reply]
@DaxServer: Wikidata is about structured data so a text like "Gift of Henry U. and Emma B. Halsted, 1968" should be converted in the relevant provenance statements. That is not an easy task at all! Multichill (talk) 16:23, 24 May 2024 (UTC)[reply]
Now I understand why. Thanks again Multichill! DaxServer (talk) 17:25, 24 May 2024 (UTC)[reply]
  • @DaxServer: Ah, now I get it. I have seen that error message before. Is there anyway the error message could be reworded to say "this property is to be used exclusively for structured data at Commons" instead of how it is currently worded? Is there a specific reason it is not allowed at Wikidata? Does it have to do with copyright of the attribution sentence? --RAN (talk) 17:01, 22 May 2024 (UTC)[reply]

January 1, 1900 versus 1900[edit]

Is there a way we can identify where we have a full date and year only date and automatically give a "preferred rank" to the full date? I may have brought it up before, but I am still processing them by hand when I identify them. RAN (talk) 05:23, 22 May 2024 (UTC)[reply]

I think it is done, with some caveats, by Wikidata:Requests for permissions/Bot/BorkedBot 5 (@BrokenSegue:). Epìdosis 05:48, 22 May 2024 (UTC)[reply]

Scottish Index of Multiple Deprivation and similar deprivation indexes[edit]

Dear all

I was wondering if there is an existing mechanism in Wikidata for adding datasets connected to deprivation indexes, such as the Scottish Index of Multiple Deprivation (SIMD) and POLAR in England?

This data is currently available under a OGL3.0 and provides rich population-level data on access to healthcare, education, income levels, rurality etc. It is comprised of datazones that are ranked approximately every 4 years and are used heavily by Scottish Government and local authorities to decide on policy changes. It is currently available for 2012, 2016 and 2020, with the next one in 2026, in an XML format. More info about 2020, including the type of data that can be downloaded, are available here.


Any thoughts about whether it is suitable for adding to Wikidata, or whether it would be better as a standalone Wikibase would be deeply appreciated. Drkirstyross (talk) 10:58, 22 May 2024 (UTC)[reply]

I don't think in principle including them would be a problem, similar to any other demographic data, but there's a major practical issue: what items would we put these on? The published data zones are MSOA or similar, very small areas - and we don't tend to have items corresponding to those geographies. We'd have to duplicate that hierarchy and then tie them back into existing regions, which would be tricky. Andrew Gray (talk) 13:17, 22 May 2024 (UTC)[reply]
That's the tricky thing. Some areas are small (a couple of streets in some cases) but others are vast (those in the Highlands and Islands) as they are designed to encompass ~500 members of the population. Each datazone can contain multiple postcodes and/or council authorities. The ranking from each year is one thing, but it is made up of a plethora of measures that might be really useful additional information.
In the interests of transparency, I am hoping to combine various open population datasets as part of an undergraduate degree module I am coordinating in September 2023, so it would be amazing if the data were incorporated into Wikidata. It won't be the end of the world if it couldn't be done and/or done by then. Apologies, not really answering your question Andrew! Just wanted to start a conversation about it, as I imagine that there are similar governmental datasets available elsewhere. Drkirstyross (talk) 15:01, 22 May 2024 (UTC)[reply]
I think if you're adding more data to these that sounds fine - my concern would be that if we just had an item saying "this tiny bit of Cumbernauld has SIMD 7, and it's in Cumbernauld", there wouldn't really be much more you could do with it. I think ONS provide enough existing data to link them back up to the overlying areas?
It might be worth running up one or two examples and seeing what they look like. Andrew Gray (talk) 23:08, 23 May 2024 (UTC)[reply]

Q11639308 and Q56651571[edit]

decommissioned (Q11639308) and out of service (Q56651571) are two terms used to describe the state of use (P5817) property. In my opinion, there is no difference between these items, so I propose to merge them. If this difference exists, just I don’t see it, please point out what it is, to make it possible to note this fact in the description of the items (especially in the case of out of service (Q56651571), the description is now meaningless). I think the items also should have the different from (P1889) property added then (as in not in operation (Q111802839)). Pyrlandczyk (talk) 08:19, 23 May 2024 (UTC)[reply]

I think something can fall out of use without being actively decommissioned — Martin (MSGJ · talk) 12:28, 23 May 2024 (UTC)[reply]
A mothballed power station can be out of service, but can return to service quite quickly. A decommissioned one normally has irrevocable changes. Vicarage (talk) 14:38, 23 May 2024 (UTC)[reply]
  • A vending machine can be temporarily "out of service" because it is empty or unplugged, but at the point where it is beyond repair it becomes "decommissioned". --RAN (talk) 16:22, 23 May 2024 (UTC)[reply]

perpetrator of holocaust[edit]

Hi everyone,

the statement perpetrator of Holocaust is not added anywhere.

Here is the fitting SPARQL statement 

SELECT DISTINCT ?item ?itemLabel WHERE {
SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE]". }
 {
   SELECT DISTINCT ?item WHERE {
     ?item p:P8031 ?statement0.
     ?statement0 (ps:P8031/(wdt:P279*)) wd:Q113546439.
   }
   LIMIT 100
 } 
 }

There is a Wikipedia category that lists those people. How do I go about batch adding those statements?

I can pipe together the various api's , but maybe there is a better way. also, if there is some kind of sandbox api endpoint, I'd be very interested as not to break anything.

Is this the right endpoint for adding a claim?


lastly, the Adolf Hitler entity is protected, so can an admin please ad purpotrator of holocaust to that enity?


best wishes. FerhatUnvar (talk) 08:29, 23 May 2024 (UTC)[reply]

The Holocaust (Q113546439) doesn't seem to be the correct item. I also think these kind of statements are too general. Sjoerd de Bruin (talk) 09:30, 23 May 2024 (UTC)[reply]
We don't have a "perpetrator of" property in Wikidata. Before you want to add anything in bulk, it would be good to understand how Wikidata works and that goes better if you add individual statements.
Q113546439 was just a doublicate that existed because nobody over at WikiBooks added it to the correct Wikidata item which is The Holocaust (Q2763) (you could easily find it by typing Holocaust into Wikipedia to see where that article is). That item actually has a statement that Hitler was a perpetrator of the Holocaust. That statement lacks citations and we have a constraint that citation are necessary for this property. ChristianKl ❪✉❫ 11:01, 23 May 2024 (UTC)[reply]
@ChristianKl
That is confusing to me as I have built that query using Wikidata Query Builder, I was trusting that the autocomplete to give me correct suggestions. I assume you now added a redirect from Q113546439 to Q2763, if so, cheers for that.
I was talking about the perpetrator property here and added "of" to make it a correct English sentence. I did add the perpetrator statement for Goebbels, so the linked statement now returns one result. I did look at the Wikipedia article for holocaust, but failed to find a reference to Q2763 or Wikidata on it, I see now it's hidden behind the "tool" dropdown. The citations constraint does not link a citation policy directly, adding a link to Help:Sources in the description seems like a good idea to me - or is there a more fitting page?. I am still unsure if linking a given person's Wikipedia article via imported from Wikimedia project (P143) meets the citation constraint. I suspect that there are a lot of wikipedia categories similar to Category:Holocaust_perpetrators that allow to infer statements, so implementing tooling to facilitate seem useful to me. My initial question is aimed if such tooling already exists and if not, how I can help in providing it. FerhatUnvar (talk) 12:16, 23 May 2024 (UTC)[reply]
Autocomplete is going to give you an item that matches the name that you are searching for. Wikidata does not yet process mind-reading techonology to know which item you actually want.
Help:Sources is unfortuantely no policy page and does not reflect current policy in regarding sources in Wikidata.
perpetrator (Q4445088) is an item and no property. The Holocaust did not perpetrate Joseph Göbbels, so the claim you added is wrong.
Claiming that someone is a Holocaust perpetrator is a quite serious charge and thus not something that shoudl be added simply because of a Wikipedia category but if such a claim would be added it should have a proper source. ChristianKl ❪✉❫ 12:37, 23 May 2024 (UTC)[reply]
Autocomplete is going to give you an item that matches the name that you are searching for. Wikidata does not yet process mind-reading techonology to know which item you actually want.
sometimes it is better just to say nothing if you have nothing constructive to say. Just feedback as you are part of the welcome new user initiative. You made me stop contributing all together.
FerhatUnvar (talk) 13:18, 23 May 2024 (UTC)[reply]
I explicitely told you that we have no "perpetrator of" property and you still assumed that another property does what you want a "perpetrator of" property to do. If you contribute on the basis of adding a lot of wrong statements to Wikidata because you don't engage in effort to understand what you are doing, that's bad for Wikidata and that's why I was as explicit as I was. ChristianKl ❪✉❫ 23:33, 23 May 2024 (UTC)[reply]
P143 is not adequate as a source for statements which are covered by the Living people policy which was mostly written by ChristianKl by the way. Sources for things like perpetrator are mandatory for living people or recently deceased. There might be people alive that was involved in the Holocaust. It would be a really bad idea to add such statements unsourced in batch. Infrastruktur (talk) 15:37, 23 May 2024 (UTC)[reply]
Relating to this I'd like to point out that we are using subject has role (P2868) for pointing people that are Holocaust victim (Q5883980) on over 128000 items. Maybe P2868 could be used for Holocaust perpetrators as well if there aren't more suitable properties? –Samoasambia ✎ 11:37, 23 May 2024 (UTC)[reply]
That usage is already a bad attempt to circumvent our decisions about property creation. Using it for perpetrators would however even worse because mistaken claims about victims are not as serious as mistaken claims about perpetrators. ChristianKl ❪✉❫ 12:39, 23 May 2024 (UTC)[reply]
I want to add statements that X was/is a perpetrator of the Holocaust, where X is a person. The Holocaust is well researched, so sources are not a problem.
Could anyone kindly point me to resources so I can add those statements in bulk (with sources) via the API? FerhatUnvar (talk) 17:59, 24 May 2024 (UTC)[reply]
You can look at comparable items and check how they do it, for example, Osama bin Laden (Q1317): significant event (P793), Khalid Sheikh Mohammed (Q319068): nothing, Hamza al-Ghamdi (Q653287)}: nothing, Mohamed Atta (Q110067): participant in (P1344) etc.
Property participant in (P1344) looks like a good candidate to me. Difool (talk) 02:00, 25 May 2024 (UTC)[reply]
the holocaust Q2763 has the P8031 perpetrator property with the Value Adolf Hitler. When I understand the definition of P8031
it should be the statement Hitler participant in (P1344) Holocaust and participant should have qualifier "object has role perpetrator" and that claim should have a citation FerhatUnvar (talk) 05:31, 25 May 2024 (UTC)[reply]
This query shows humans, with a statement subject has role (P2868) perpetrator (Q4445088):
SELECT ?human ?humanLabel ?y WHERE {
  ?human wdt:P31 wd:Q5.
  ?human ?x ?y.
  ?y pq:P2868 wd:Q4445088.
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en". }
}
Try it!
You could also use convicted of (P1399), probably less controversial. Difool (talk) 08:38, 25 May 2024 (UTC)[reply]
Still objecting mass-adding statements like this without proper sourcing. Sjoerd de Bruin (talk) 08:47, 25 May 2024 (UTC)[reply]
Yes, I would call it a requirement, but I assumed the statements would be added with sources after reading "sources are not a problem" above. Difool (talk) 09:28, 25 May 2024 (UTC)[reply]
@Sjoerddebruin I plan add sources as @Difool suspected. I also want to follow all rules and requirements. That's the whole reason I am writing here instead of just doing it. I don't see a reason to repeatedly do the same manual steps if I can write a script once, create a table containing person ids and sources , to then finally run the script. FerhatUnvar (talk) 11:33, 25 May 2024 (UTC)[reply]
Thank you @Difoolfor the query, but what I am looking for would be an insert statement/api call.
There are several people that perpetrated the holocaust but never were convicted. Also, holocaust is not a criminal charge. I could however add the charges from the Nuremberg trials. FerhatUnvar (talk) 11:38, 25 May 2024 (UTC)[reply]
It's not trivial to decide who was a perpetrator of the Holocaust. Were the perpetrators only those people who actively killed Jews? Those who helped to kill them? Those who commanded killing them? Or everybody who was a member of the Nazi Party? You definitely need a good source to add such statements. What soure are you planning to use? D3rT!m (talk) 11:51, 25 May 2024 (UTC)[reply]
I will add undisputed perpetrators. This is not a fact finding commission. I will follow the guidelines that apply to Wikipedia, meaning to stick to the scientific consensus and sources that satisfy academic criteria. FerhatUnvar (talk) 12:13, 25 May 2024 (UTC)[reply]
@FerhatUnvar You can use the query to see how the statements are used, i.e. not a lot, and mostly for murders. I would use QuickStatements to add the statements. Difool (talk) 12:22, 25 May 2024 (UTC)[reply]

Conflation[edit]

I created Zig Zag Productions (Q126074734) because the VIAF currently attached was being used by 4 completely different entries, a rolling paper, a magazine a Japanese band and a Latvian band. If someone can figure out the nationality of the musical group, it can be changed from "conflation" to "musical group", or merged with an existing band. I have not been able to figure it out. RAN (talk) 16:26, 23 May 2024 (UTC)[reply]

Ok it does not make sense f Elonathi (talk) 19:10, 23 May 2024 (UTC)[reply]
If it's the list of works in https://viaf.org/viaf/127976276/ it seems to be related to https://www.discogs.com/label/215164-Zig-Zag-Productions and https://www.discogs.com/label/2736131-Zig-Zag-Productions-USA - I looked at a few and they seem to be mentioned in the copyright information rather than a specific role. It seems to be a company, and is mentioned in https://books.google.co.uk/books?id=myxMAAAAYAAJ&focus=searchwithinvolume&q=zig+zag+productions where it says Christian De Walden (Q5109486) is president and producer. What I'm unsure of is if that and https://catalogue.bnf.fr/ark:/12148/cb138847893 (the source of the list of works in VIAF) are the same company, or if there are separate Italian and American companies both associated with De Walden. Peter James (talk) 19:12, 23 May 2024 (UTC)[reply]
Most of the records listed at https://d-nb.info/gnd/10285190-6 seem to be by https://www.discogs.com/artist/1048919-Zig-Zag-6 which is a pop rap group, possibly Dutch or German, active in the 1990s, but one is https://www.discogs.com/release/194092-Various-Stickman-Musica where "Zig Zag" is https://www.discogs.com/artist/662191-Zig-Zag-3 who is Canadian. Peter James (talk) 19:22, 23 May 2024 (UTC)[reply]
http://isni.org/isni/0000000122539320 has "Italian" from the the BNF, the list of records from DNB, and a link to http://musicbrainz.org/artist/5d62061d-73ed-4478-b5d2-dcc6792ae857 (also https://www.discogs.com/artist/617824-Zig-Zag-2) which is a synth-pop group, probably German, active in the 1980s. Peter James (talk) 19:31, 23 May 2024 (UTC)[reply]
The ISNI website has a form on each ISNI page to report errors, sometimes they create new ISNI when a conflation in an ISNI record is reported and one of the conflated entities doesn't yet have an ISNI. ISNIplus (talk) 06:22, 24 May 2024 (UTC)[reply]
Several initial ISNI records were created from VIAF clusters, if VIAF had a conflation at that time it could end up in ISNI. ISNIplus (talk) 06:32, 24 May 2024 (UTC)[reply]
  • Any solutions? Should it be kept as an conflation or converted to Zig Zag Productions, I will make it Zig Zag Productions for now, you can move it back if disagree. --RAN (talk) 00:58, 24 May 2024 (UTC)[reply]

Charles Edwin Odegaard (Q5081342)[edit]

I've deprecated the image, because it is apparently the wrong Charles Odegaard. Should it instead be removed entirely? Because we do not have any correct image of him, its inclusion (even deprecated) makes it appear in Infoboxes, etc. - Jmabel (talk) 21:40, 23 May 2024 (UTC)[reply]

This is a good thing. Errors that are visible are more likely to be fixed. Hiding the problem doesn't fix anything. It falls on Hungarian Wikipedia to make sure their infoboxes are properly programmed. Infrastruktur (talk) 22:29, 23 May 2024 (UTC)[reply]
I can't speak for every infobox on every project, but generally they should only be using "truthy" statements, so deprecating the claim will make it not used. One advantage to deprecating over deletion is that a bot will not attempt to re-add a deprecated claim. Bovlb (talk) 22:59, 23 May 2024 (UTC)[reply]
  • Jmabel, amazing detective work. I created an entry for the other person and migrated the image there, still deprecated until we get more evidence, I found his obit, and now looking for an image in newspapers from when he was appointed. The Flickr account never misidentifies him, it looks like we misidentified him. --RAN (talk) 02:45, 24 May 2024 (UTC)[reply]
"Mr. Odegaard is not to be confused with, and is not related to, Charles E. Odegaard, former University of Washington president, though the two were good friends. Evelyn Odegaard said they both took to signing their names with their middle initials to avoid confusion. They also had an unofficial pact to both do good work to better their shared name, she said." --RAN (talk) 02:48, 24 May 2024 (UTC)[reply]

Week passed since 2024-05-16 without new Wikidata:Database reports/Constraint violations[edit]

Special:Contributions/KrBot2 ISNIplus (talk) 06:06, 24 May 2024 (UTC)[reply]

@ISNIplus: that bot has an operator. If a bot is not functioning as it supposed to, you should always contact the operator. No point in complaining here at all.
This bot runs on private infrastructure on closed source code so you really need Ivan to fix it. Multichill (talk) 16:09, 24 May 2024 (UTC)[reply]

these two pages (on English and Arabic wiki) should be connected, but it won't work[edit]

I've tried to connect these two pages, and it seems to be a bit complicated.

renamed from en:2014 Nahal Oz military base raid

these pages here

I think I need to merge those, but I don't know how?

FourPi (talk) 10:04, 24 May 2024 (UTC)[reply]

Hey, thanks for noting us. I have now merged the items into 2014 Nahal Oz attack (Q20420465). Next time you can find instructions at Help:Merge. –Samoasambia ✎ 10:17, 24 May 2024 (UTC)[reply]
@Samoasambia thank you. FourPi (talk) 10:20, 24 May 2024 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat&oldid=2164226933"