User talk:Qono

From Wikidata
Jump to navigation Jump to search
Unblock request granted

This blocked user asked to be unblocked, and one or more administrators has reviewed and granted this request.

Qono
block log • ipblocklist • crossblock • luxo's • unblock • remove gblock • contribs: +/-
Request reason:
My block is unwarranted. My edits are constructive and well-sourced and I responded to concerns about the edits on my talk page, pointing to Wikidata guidance. The administrator blocked me despite being unable to get support for action against me on the administrator's noticeboard. The administrator seems to not understand that there can be multiple claims about a property from multiple sources. They also have ideas about which values are valid for a property that contradicts the value type constraints of those properties.
Unblock reason:
Conversation is happening. Assuming good faith. Multichill (talk) 16:55, 6 October 2020 (UTC)[reply]


This template should be archived normally.

беларуская (тарашкевіца) | български | čeština | English | español | français | македонски | português | português do Brasil | русский | svenska | 中文 | +/−

Continuing archived discussion[edit]

@Fuzheado, Multichill: I can't respond in the original thread, so I am responding here.

I'm happy to field questions about my edits and did engage with the editors in the earlier thread. I also largely paused my editing to hear from other admins on the noticeboard post that Multichill made. I agree that the site you mentioned that I used as a source is not the most authoritative, but the information there is largely good and my edits attribute those statements to that source. I believe those edits are adding useful, sourced information that is missing from most all of the items I've updated and I hope that my edits will make it easier for contributors to match items to those more authoritative sources. Those more authoritative sources should certainly be given priority when those statements are added.

Apologies for characterizing the block as harassment. I am making good faith edits and have engaged with those with questions, so the block seemed excessive to me, but I will strike that comment when I am unblocked and able to. Qono (talk) 21:00, 5 October 2020 (UTC)[reply]

Hello,
Yes the data is interesting for the project but there are too many problems in those editions. The blocking is justified. You have been warranted of precedent issues and did not reply.
One of the main problem is that you make mass edits without knowing well the editorial logic of wikidata.
Before doing mass uploads, you should more practice manual edition and explore the items.
For example of issues, a query on genre (P136) edits that you made in which the value is an item with instance of (P31) art movement (Q968159) that should be associtated with property movement (P135).
SELECT ?item ?itemLabel ?genreLabel 
WHERE 
{
  ?item p:P136 ?genredeclaration.
  ?genredeclaration ps:P136 ?genre.
  ?genre wdt:P31 wd:Q968159.
  MINUS {?genre wdt:P31 wd:Q1792379.}
  
  ?genredeclaration prov:wasDerivedFrom ?ref.
  ?ref pr:P854 ?URLref.
  FILTER regex(str(?URLref), "beckchris.wordpress.com")

  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
}
Try it!
All those statements need to be fixed. A massive edition should not imply a massive cleanup by other contributors.
The other issue is the data quality. As already pointed in the discussion page, in your edits there are claims that are obviously wrong without controversy (example: location (P276) Spain (Q29) in Les Demoiselles d'Avignon (Q910199), https://www.wikidata.org/w/index.php?title=Q910199&diff=1275998891&oldid=1237071024 ) and the nature of the reference, a blog, is an issue especially because there is no source in this reference. There are usefull information on those pages but they need to be more carefully checked before being reused. Best regards --Shonagon (talk) 08:48, 6 October 2020 (UTC)[reply]
Hi, @Shonagon: I did engage with editors, as is evident in the earlier discussion. Regarding the problem you illustrate with your SPARQL query, those issues only arose after the administrator that blocked me changed the value type constraint for genre (P136) to remove art movement (Q968159), seemingly to retroactively justify the rationale for the block after I pointed out why the edits were valid in the discussion at the earlier thread. At the time that I made the edits, all of the edits were valid. OpenRefine flags any potential issues like this, and I was careful to correct any issues before posting. Finally, I concede that the source is not ideal. As you said, I do think it provides useful information, but I can agree to not use it in mass edits in the future. I think the issue with location in those edits is due to the source sometimes referring to the location of the work and sometimes referring to the nationality of the creator. This is a legitimate issue, but not one that I think is terribly widespread. I will be happy to clean up that issue once my block is removed. Qono (talk) 15:20, 6 October 2020 (UTC)[reply]
I merely undid your earlier mistake. I'm willing to accept good faith here, but you have to understand and acknowledge the concerns about data quality brought forward. Multichill (talk) 16:55, 6 October 2020 (UTC)[reply]
@Multichill, Shonagon: Thank you for the unblock and apologies for misunderstanding your edit on genre (P136). I admit that I mixed up subject type constraint (Q21503250) and value-type constraint (Q21510865). Your reverting my change on that property was correct. Given that, and based on Shonagan's SPARQL query, I'm not seeing any constraint violations for genre (Q483394) on my edits. I welcome another SPARQL query that lists any violations on my edits. That would help me clean up any entries that need it. Qono (talk) 14:01, 7 October 2020 (UTC)[reply]
Hello Qono,
Unfortunatly, you do not seem to have understood the issue. It is an editorial issue. Constraints of properties are indicative and not strict as datatypes, because there are often limit cases. Here there is no ambiguity:
Expressionism (Q80113) is a movment art movement (Q968159) ( https://w.wiki/g9u ), not a art genre (Q1792379) ( https://w.wiki/g9y )
Impressionism (Q40415) is a movment art movement (Q968159) ( https://w.wiki/gAH), not a art genre (Q1792379) ( https://w.wiki/gAJ )
futurism (Q131221) is a movment art movement (Q968159) ( https://w.wiki/gA4 ), not a art genre (Q1792379) ( https://w.wiki/gA3)
Contributors do sometimes mistakes on this with manual edition. Here, with massive edition, you put a lot of bad edits. Now a large proportion of genre (P136) bad claims on visual artworks come frome your edits.
This a a serious issue because those information are reused, including in other Wikimedia projects like Wikipedia. For example: Le Déjeuner sur l'herbe on Wikipedia FR. It has a significant impact on the quality of data coming from Wikidata and we must pay attention on this.
However, almost all of the items listed in the query have already appropriated genres provided by others contributors. Your massive edit put here a mess with inappropriate values. Once again, there are editorial logic and practice in Wikidata that you don't know well.
All of these genre (P136) edits by you listed in SparqL's result should be removed. Will you do it? Best regards Shonagon (talk) 09:57, 10 October 2020 (UTC)[reply]
@Shonagon: Thank you for clarifying your concern. I don't think that the structure and statements in Wikidata align with your stricter understanding of what counts as a genre. In the examples you've given, Impressionism (Q40415) is listed as an instance of (P31) genre (Q483394) and Expressionism (Q80113) is listed as an instance of (P31) art style (Q1792644), which is a subclass of (P279) art genre (Q1792379). Therefore, it seems sensible to me to include them in genre (P136) statements for works. And, again, art movement (Q968159) is a valid value for genre (P136), which makes futurism (Q131221) a valid statement as well.
Since you seem to come from an art history background, I can understand why you would define genre narrowly in the sense of genre of painting (Q16743958), but it seems to me that the Wikidata property is meant to capture a broader sense that might be better labeled "genre or style". This is reinforced by the fact that genre (P136) is the only way (that I can see) to ascribe a style to a work.
If you would like to continue this discussion, I think the talk page for genre (P136) would be the best place so that we can involve more editors. If the consensus is that art movement (Q968159) should not be included as a value type constraint for genre (P136), then I will happily remove the statements that you point out that would then become invalid. You could also put forth a property proposal for "art genre" that is more strictly constrained in the way that you think is appropriate. Cheers, Qono (talk) 02:13, 12 October 2020 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=User_talk:Qono&oldid=1289994258"