Wikidata:Events/Data Modelling Days 2023/Clinic2
✨---------------✨---------------✨---------------✨---------------✨---------------
Data Modelling Clinic #2
Facilitators: Lydia Pintscher (WMDE), Arian Bozorg (WMDE)
✨---------------✨---------------✨---------------✨---------------✨---------------
Previous: https://etherpad.wikimedia.org/p/DMD2023-FeedbackALG
Next: TBD
👥 Number of participants (including speakers): ...
51 (at 00:20)
🖊️ Notes & links
Hypotheses, theories, lemmas, theorems and similar concepts (Daniel Mietchen)
...
❓ Questions and discussions
- copy paste**
- copy paste**
Username:
My modelling challenge (approximately 3 sentences):
My possible solutions:
I am interested:
I can help:
Username: Aap1890
My modelling challenge:
I’m adding items for publications by two art historians I study. I am very interested in how they signed their work (with pseudonyms, etc). Some works, however, are unsigned or I don’t know how they were signed, and after I uploaded my data I realized I have a modeling problem and I made a mess. Note: I will be changing subject named as qualifiers to object named as...that was also confusing to me, but I think I get it now :)
Here are some examples:
Unsigned: https://w.wiki/8L7y
Unknown (where I haven’t gotten a copy of the article yet so I don’t know how it was signed): https://w.wiki/8L8C
Follow-up: i guess one follow up question i have is how to add unknown or no value using openrefine or quickstatements. if anyone knows that, please follow up with me!
somevalue for unknown value
novalue for no value
Both novalue and somevalue should be used without double quotes around it.
https://www.wikidata.org/wiki/Help:QuickStatements#Samples_by_value_type
My possible solutions:
Should I set to unknown value, a la what WikiProject Sum of All Paintings does? https://www.wikidata.org/wiki/Wikidata:WikiProject_sum_of_all_paintings/Anonymous_creator
Or novalue?
I am interested:
I can help:
Username: Fjjulien
My modelling challenge:
Relationships between ethnic groups are inconsitently modelled. At a high level, "narrower" type relationships with Indigenous people (Q103817) are modelled with subclass of (P279). At more granular levels, we find a mix of P279 and part of (P361) relationship. Which one is the most appropriate modelling strategy?
See this slide from WikidataCon 2021: https://commons.wikimedia.org/w/index.php?title=File%3AWikidataCon_2021_Indigenous_knowledge.pdf&page=12
This other slide includes queries using combinations of P279 and P361: https://docs.google.com/presentation/d/1cV2Ey23UtaglkL4maAb89vXnMvuhEEGF5vjl9KuyG-k/edit#slide=id.g22d08d52285_0_24
My possible solutions:
The discussion can be continued via this WikiProject:
https://www.wikidata.org/wiki/Wikidata:WikiProject_Indigenous_peoples_of_North_America
I am interested:
jshieh
jshieh says:I'd be interested in participating in this discussion!
Mairelys
Mairelys says:I'm interested as well!
jshieh
jshieh says:It is also relevant of other ethnic groups, historical ethnic groups that no longer exist
Susanna Ånäs
Susanna Ånäs says:I think it would be great to keep the different scopes in check
I can help:
Username: Andy Mabbett; Pigsonthewing
My modelling challenge:
How can we improve oniscophagy (the act of eating woodlice (pillbugs; Oniscoidea);
Q123579031) and oniscophage (organism that eats woodlice; Q123579039), and the modelling of the relationships between them, and with Oniscoidea (Q7093945)?
Do we have other "-phagy" items that can serve as a template?
necrophage (Q25524623) / necrophagia (Q2961137)
Useful, thank you, I made some changes based on those, but there is more to do
My possible solutions:
I am interested:
I can help:
Username: Ciell
My modelling challenge:
I have mapped structures for Wikidata items (by example of the Sum of all paintings), but how do I make others aware of this mapping so they don't 'destroy' the work done because they are unaware of the larger mapping? What is the best way of communition of mappings and models?
https://www.wikidata.org/wiki/Wikidata:WikiProject_Witches/Properties
https://www.wikidata.org/wiki/User:Ciell/Sexual_%26_reproductive_health#Wikidata:WikiProject_Sexual_&_reproductive_health/Properties
My possible solutions:
Is there a template to put on the talkpage (or item)?
Is there an property for this?
Other?
How about setting up a Wikiproject (or sub-project)?
Question remains though! How can I make people aware about the mapping laid out in WikiProject Witches? And even more with the Sexual and Reproductive Health, that was actually "completed" but gets changed by others still (AGF: unaware of the logic behind the mapping/model).
on focus list of Wikimedia project (P5008)
maintained by WikiProject (P6104)
(there was discussion of merging those properties: https://www.wikidata.org/wiki/Wikidata_talk:WikiProjects#Consistency_in_tagging_WikiProjects_via_P5008_and/_or_P6104 )
Thanks - not sure if people notice that though, esp. with the sexual health people don't seem to first scroll through all properties before they start 'correcting'.
One thing you might consider is automatic rewrites that "fix" incorrect edits.
I am interested:
I can help:
Username: Peter Patel-Schneider (pfps)
My modelling challenge:
I am trying to figure out how to move forward in modeling disease. The problem is that there are two different mechanisms currently in place.
On one side is disease (Q12136) and on the other is class of disease (Q112193867) and disease of a particular individual (Q112193769). Disease (Q12136) is the obvious class for diseases but it is ontologically diverse, with named diseases as both instances and subclasses. Class of disease (Q112193867) and disease of a particular individual (Q112193769) were created, I think, to be a better organization of diseases.
One aspect of the problem is that it is unclear how to actually model diseases. Should there be an instance-type distinction between "single" diseases like measles (Q79793) and "complex" diseases like cancer (Q12078). Or should this be a superclass/subclass relationship. The argument for the latter is that it is very hard to determine exactly what a simple disease is. For example, should the line be drawn at the level of a individual infectious agents, creating chicken pox and shingles as a single disease. Or should the line be drawn at the level of symptoms, separating chicken pox and shingles but also separating measles that does not cause sterility and measles that does cause sterility or separating juvenile measles and adult measles.
My possible solutions:
I would prefer using only class of disease and a generalization relationship for diseases, with the individuals being (only) disease of particular individual. But some argue that
I am interested:
I can help:
Username: Jheald
(copied forward to next session)
My modelling challenge:
Examples of classes that are subclasses of *both* "abstract entity" (Q7048977) and "material entity" Q53617407
https://query.wikidata.org/embed.html#SELECT%20DISTINCT%20%3Fclass1%20%3Fclass1Label%20%3FphysClass%20%3FphysClassLabel%20%3FabsClass%20%3FabsClassLabel%0AWITH%20%7B%0A%20%20SELECT%20DISTINCT%20%3Fclass%20WHERE%20%7B%0A%20%20%20%20%3Fclass%20wdt%3AP279%2B%20wd%3AQ53617407%20%20.%20%0A%23%20%20%20%20hint%3APrior%20hint%3Agearing%20%22reverse%22.%0A%20%20%7D%20LIMIT%20100000%0A%7D%20AS%20%25physicalClasses%0AWITH%20%7B%0A%20%20SELECT%20DISTINCT%20%3Fclass%20WHERE%20%7B%0A%20%20%20%20%3Fclass%20wdt%3AP279%2B%20%20wd%3AQ7048977%20.%20%0A%23%20%20%20%20hint%3APrior%20hint%3Agearing%20%22reverse%22.%0A%20%20%7D%20LIMIT%20100000%0A%7D%20AS%20%25abstractClasses%0AWITH%20%7B%0A%20%20%20SELECT%20DISTINCT%20%3Fclass%20WHERE%20%7B%0A%20%20%20%20%20%20INCLUDE%20%25physicalClasses%20.%0A%20%20%20%20%20%20INCLUDE%20%25abstractClasses%20.%0A%20%20%20%7D%0A%7D%20AS%20%25commonClasses%20%0AWITH%20%7B%0A%20%20%20SELECT%20DISTINCT%20%3Fclass%20WHERE%20%7B%0A%20%20%20%20%20INCLUDE%20%25commonClasses.%0A%20%20%20%20%20MINUS%20%7B%0A%20%20%20%20%20%20%20%7B%20SELECT%20%28%3Fclass%20AS%20%3FparentClass%29%20WHERE%20%7B%20INCLUDE%20%25commonClasses.%20%7D%20%7D%0A%20%20%20%20%20%20%20%3Fclass%20wdt%3AP279%20%3FparentClass.%0A%20%20%20%20%20%7D%0A%20%20%20%7D%0A%7D%20AS%20%25reducedClasses%0AWHERE%20%7B%0A%20%20INCLUDE%20%25reducedClasses%20.%0A%20%20%3Fclass%20wdt%3AP279%2a%20%3Fclass1%20.%0A%20%20hint%3APrior%20hint%3Agearing%20%22forward%22.%0A%20%20%3Fclass1%20wdt%3AP279%20%3FabsClass%20.%0A%20%20%3FabsClass%20wdt%3AP279%2a%20wd%3AQ7048977%20.%0A%20%20hint%3APrior%20hint%3Agearing%20%22forward%22.%0A%20%20OPTIONAL%20%20%7B%0A%20%20%20%20%20%20%3FabsClass%20wdt%3AP279%2a%20wd%3AQ223557%20.%0A%20%20%20%20%20%20hint%3APrior%20hint%3Agearing%20%22forward%22.%0A%20%20%20%20%20%20BIND%20%281%20AS%20%3Fno_go%29%20.%0A%20%20%7D%0A%20%20FILTER%20%28%21bound%28%3Fno_go%29%29%20.%0A%20%20%3Fclass1%20wdt%3AP279%20%3FphysClass%20.%0A%20%20%3FphysClass%20wdt%3AP279%2a%20wd%3AQ223557%20.%0A%20%20hint%3APrior%20hint%3Agearing%20%22forward%22.%0A%23%20%20OPTIONAL%20%7B%0A%23%20%20%20%20%20%20%3FphysClass%20wdt%3AP279%2a%20wd%3AQ7184903%20.%0A%23%20%20%20%20%20%20hint%3APrior%20hint%3Agearing%20%22forward%22.%0A%23%20%20%20%20%20%20BIND%20%281%20AS%20%3Fno_go2%29%20.%0A%23%20%20%7D%20%20%20%20%0A%23%20%20FILTER%20%28%21bound%28%3Fno_go2%29%29%20.%0A%20%20SERVICE%20wikibase%3Alabel%20%7B%20bd%3AserviceParam%20wikibase%3Alanguage%20%22%5BAUTO_LANGUAGE%5D%2Cen%22.%20%7D%0A%7D%20ORDER%20BY%20lcase%28%3FabsClassLabel%29%0A
My possible solutions:
This may be a reflection of two different ways we think of a class -- as a collection of things that are concrete; but also as a container of things (a container which itself is abstract).
We need to have a clearer think about this, because being able to say whether the contents of a class are material things or are abstract things is a really valuable thing to be able to say
I am interested:
I can help:
Username:
My modelling challenge (approximately 3 sentences):
Hypotheses, theories, lemmas, theorems and similar concepts
My possible solutions:
are being explored at https://ecology-22.wikibase.cloud/wiki/Main_Page#Example_queries
I am interested:
I can help:
🎯 Key takeaways and outcomes
...
...
☑️ Next steps
...
...