Hoi,
I cannot agree with you at all on this. The main point here is that it is save for projects like the BLT to use Wikidata to set up the data for the people they deem to be notable. In this, notable on a Wikipedia level. So the point is to build the list find the sources etc. Now this whole point about enforcing Wikidata notability means that there is no room to grow. Every John, Dick and Harry can come along (and usually does) to nitpick and remove items. This destroys the integrity of the accumulated data. It destroys Wikidata as a tool to bring diversity to the Wikimedia projects.

To be brutal. A friend lives in a town where many Yazidi refugees live. I have considered Wikidata as a staging project for them  but because I do not consider Wikidata to be safe for staging data for them, I did not even suggest it. Consequently I do not help them build on the idea to document their culture. If you do not know about Yazidis.. shame on you.
Thanks,
       GerardM

On 8 January 2018 at 04:58, Charles Horn <charles.horn@gmail.com> wrote:


On 7 January 2018 at 04:22, Brill Lyle <wp.brilllyle@gmail.com> wrote:
Hi Charles,

Thanks for taking the time to try to understand the issues that are being raised here.
 
- Wikipedia initiatives need a unique identifier in Wikidata they can use on Wikidata to tag items to their initiative -- and most importantly run SPARQL queries on
-- I don't think a Q number will work for this purpose, though I'm not sure
-- Whatever solution will allow for this functionality that the community can come to consensus on would be used
--- We are not tied to catalog in any way except it has worked as a solution for our scenario
--- Catalog was suggested to us with consensus. But if there's a better option that meets this need, we will use that one
- Wikipedia initiatives can then add location and date to the SPARQL query for specific events that can generate an event-based task list Listeria table
- Wikipedia can also run SPARQL queries by geographic data (place of birth, residence, place of death, etc.) to find, within its tagged items, a suggested list of pages to work on for that geographic location
- Wikidata is key because it is possible to create Wikidata items that don't yet exist on the various language Wikipedia. Wikidata allows organizers to create a scaffolding where notability is the ongoing, over-arching goal, so that new Wikipedia pages in various languages based on this Wikidata scaffolding can be easily created. This also makes Wikidata part of every single outreach event, which to me seems a logical, forward-moving, innovative thing, as this does not typically happen at most editathons. If I see librarians at editathons I "target" them specifically because they typically understand identifiers, Authority control, and the value of VIAF. This again is done to establish and improve notability. I would assume that a significant portion of my 10,000 manual edits on Wikidata are identifiers that I pull from VIAF.


Thanks Erika, this is a nice summary of the requirements which I hope will be useful for others as well. I feel I understand and accept that the goal of Wikipedia initiatives such as BLT is to correctly demonstrate notability in Wikidata, and subsequently utilise this to allow the creation of good quality articles about notable individuals and topics, in multiple languages, on Wikipedia. I think this is a worthy and reasonable goal, and innovative as you say, so am happy to support it, and I sincerely hope there is a way forward here, to get consensus, and furthers the goals of Wikidata and outreach projects.


I don't want to hijack the thread to go on about P972:catalog, because I think the discussion has already moved on to focus on the real question, which you state below: 
"All of this information is helpful in providing context. But really, the question is whether or not Wikidata is okay with integration into Wikipedia outreach. Is Wikidata a sacrosanct island of pristine metadata that is intended only for scientific scholarly research queries, trivia, etc. Or is Wikidata flexible and willing to engage with the various projects in new and exciting ways?"

but initially I thought, as others have too, that there was a reasonable and simple way forward to remove the constraint violations that triggered the bulk deletions of the property by Multichill, which to be clear I consider to be completely the wrong decision, and unfair give the short time frame, and holiday period etc (which has been said by others elsewhere). It turns out fixing the obvious constraint violations was not going to be good enough for some -- and I believe that has more to do with P972:catlog property being unclearly defined and prone to misuse, and not really applicable outside a narrow definition of catalog, rather than a general catalog concpet that it looks like it should represent. I am claiming that any abuse of catalog has more to do with P972:catalog than the outreach projects that have been using it. Perhaps I am struggling to get my point across, but it is technical, and given "We are not tied to catalog in any way except it has worked as a solution for our scenario", I accept this discussion is not the best place to make it. I was planning to do so on the wiki page... stay tuned to that if interested. The short version is that seemingly P972 as a property is supposed to infer an item, shuch as an exhibition, "owns" or "produced" a catalog, rather than is part of a catalog. This seem inconsistent with its usage as a qualifier, and of all the uses on Wikidata, two-thirds are this wiki-project usage, one-third are other "incorrect" uses of the catalog property to imply "membership of" various lists, and a small handful that is pretty much impossible to extract from the incorrect uses are "correct" uses to imply an event or institution has or produced a catalog. 

I guess a point to make against some of the attitudes here is that if people were really concerned about proper usage of P972:catalog, then more effort could have gone into making its definition clearer, and the other 1200 or so other non constraint-violating and non-outreach uses which are still apparently incorrect could have been fixed in the past. I'm happy to take up some of the effort to untangle P972:catalog for its own sake, but it feels like a different problem that the one Erika is struggling to get some consensus on and a clear way forward.

Not addressed to Erika, but to those opposed to the Wiki Project usage: I'll restate my initial point that it was difficult as someone new to the discussion, even after reading all the historical on-wiki discussion, what the exact nature of the objections were. As Gerrard summed up there were opinions, but not arguments for why. I found it difficult to agree or disagree since I could not find a clear argument or statement about what exactly was wrong, just opinions. And I did read the wiki pages too.

If anyone thinks P972 is salvageable in some way for this usage, I'm happy to try an help fix things, but right now it feels like it is an unfortunate distraction, and I was hoping a new property that has more clarity on how to correctly use it would be an acceptable way forward.



Here's an example entry, for the children's book illustrator Carole Byard
Inline image 1
The Wikipedia editathon was held at the Brooklyn Museum last summer and was connected to a museum exhibit there.

- BLT-tagged items might logically include attendees of events but typically that is not the case

I thought that the above was possible, but as you have confirmed, not the norm, and this seems perfectly reasonable to me, that attendees may happen to be notable, given all normal criteria. I didn't think anyone was claiming an individual was notable entirely because they attended an event. I spent some time searching for either Mos Def or Angel Haze attended a BLT meeting but found no evidence :)

 
- BLT-tagged items are typically curated with the assistance of the institution hosting the event, who have expertise in the local and regional artist community, and can provide BLT with a list of underrepresented visual artists from the African diaspora. It is often difficult for this population to have even local or regional coverage in press and scholarly works, so this curation is even more important. Heather and Jina (who founded BLT) will then vet this list and create Wikidata items, tagging them as BLT, location, date (to add to a Listeria task list) and begin the process of establishing notability on Wikidata in preparation for Wikipedia articles
- BLT is a two person outreach initiative made up of Heather (a visual artist) and Jina (an artist and professor). They have minimal grant funding for the Black Lunch Table project from outside sources that they list here: https://en.wikipedia.org/wiki/Wikipedia:Meetup/Black_Lunch_Table/About There is a studio assistant who is updating Wikidata items. Gerard did the first heavy lifting pass, and I have assisted with the editathon task lists. But that's it.
- Wikipedia is just one part of the BLT project


I am happy to assist the project adding VIAF for authors and musicians if that will help. 
 
- This Listeria task list model is being used in various projects
-- Plants And People (New York Botanical Garden and various Council on Botanical and Horticultural Libraries participate in this initiative): https://en.wikipedia.org/wiki/Wikipedia:Meetup/PlantsAndPeople/Lists_of_Articles

Also using this idea:
-- Women in Red

All of this information is helpful in providing context. But really, the question is whether or not Wikidata is okay with integration into Wikipedia outreach. Is Wikidata a sacrosanct island of pristine metadata that is intended only for scientific scholarly research queries, trivia, etc. Or is Wikidata flexible and willing to engage with the various projects in new and exciting ways?

- Erika



Thanks,
Charles. 

_______________________________________________
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata