search

welfare-state-analytics / riksdagen-corpus

Swedish parliamentary proceedings - Riksdagens protokoll 1867-today
Other
26 stars 5 forks source link

Fix party for top known MP but unknown party #324

Closed fredrik1984 closed 1 year ago

fredrik1984 commented 1 year ago

Based on the 0.9 corpus version, we should fix the top known MP but with an unknown party. Most of these MPs are from the 19th century. The top 100 known MPs but with an unknown party account for 22 000 hits. Adding party to each MP's Wikidata post would perhaps be a start.

@salgo60 is this something that you would like to take on?

known_person_but_unknown_party.csv

salgo60 commented 1 year ago

@fredrik1984 My focus is to connect people to the book Två kammar Riksdagen right now about 600 to do see https://github.com/salgo60/Wikidata_riksdagen-corpus/issues/38#issuecomment-1360364007

I can try to follow your ranking list when I am back home this week

unknown party

I use "no value" for people in the book that has no party set

I am in a hammock out kayaking but try some SpARQL on my mobile

salgo60 commented 1 year ago

Your 1st on the list https://www.wikidata.org/wiki/Q5616733

Was updated by me feb 2022 and then I didn't add party from the book at SPA so there is a lot of cleaning needed ...

SPA for Q5616733 https://portrattarkiv.se/details/sj9PGLAlnmUAAAAAABgdHQ

We also have a lot of parties set by other Wikidata users that is not following "the book" I.e you can filter on the timeline%20%20(sample(?bild)%20AS%20?bild)%20%0A?birth%20?death%20?partyLabel%20WHERE%20%7B%0A%0A%20%20VALUES%20?member%20%7B%0A%20%20%20%20wd:Q33071890%20%0A%20%20%20%20wd:Q81531912%20%0A%20%20%20%20%23wd:Q82697153%20%0A%20%20%20%20wd:Q10655178%20%0A%20%20%7D%0A%20%20?person%20wdt:P39%20?member.%0A%20%20OPTIONAL%7B?person%20wdt:P102%20?party%7D%0A%20%20OPTIONAL%7B?person%20wdt:P569%20?birth%7D%0A%20%20OPTIONAL%7B?person%20wdt:P570%20?death%7D%0A%0A%0A%20%20OPTIONAL%20%7B?person%20wdt:P18%20?bild%7D.%0A%20%0A%20%20BIND(URI(CONCAT(%22https://portrattarkiv.se/details/%22,?SPAid))%20AS%20?SPA)%0A%0A%20%20SERVICE%20wikibase:label%20%7B%20bd:serviceParam%20wikibase:language%20%22sv,en%22.%20%7D%0A%7D%20GROUP%20BY%20%20?person%20?personLabel%20%20?death%20?birth%20?partyLabel%0Aorder%20by%20?partyLabel&md=true&g=article&l=person&t=personLabel&s=birth&e=death&i=bild&d=0&c=partyLabel&f=partyLabel&v=t) and see parties at the wrong Time period is my feeling

image

fredrik1984 commented 1 year ago

Yes, adding party is of course a bit tricky when it comes to the 19th century. And as we have discussed previously, we use the bio books as bible. The abbreviation register at the end of the bio books, and the hierarchical tree of party formations/splits, are good sources for this work.

So in the case of Anders Danielsson (https://www.wikidata.org/wiki/Q5616733) he was a member of Lantmannapartiet (lmp) 1873–1887, Nya lantmannapartiet (nya lmp) 1884–1894), and then to the reappearing (!) Lantmannapartiet (lmp) 1895–1897.

salgo60 commented 1 year ago

Yes and in the Anders case I "connected" him to the book and didn't add party

Now when I connect to the book I

I have 11 m/s wind right now so we never know how my kayaking will end maybe we get another "Titanic submarine" news story 😅

fredrik1984 commented 1 year ago

Ok! Good luck with the kayaking!

salgo60 commented 1 year ago

Unknow parties - we need sources.....

@fredrik1984 the list you somehow has created - I guess people identified speaking in the PM

WD modulering - opolitisk - partipolistisk obunden

Jag vet inte om WD har en bra modell för hur "opolitisk...." skall hanteras se Reidunn Laurén....

salgo60 commented 1 year ago

Status: Touch and go

I have been home for some days and the following has been done,.... now I am leaving again... 1) @dpriskorn have given some care to EntityShape that I use in the below Notebook... video 1) we have since earlier played with ShEx see #129 1) WIkimedia has a hosted Notebook environment PAWS --> I created a Notebook "Test validate people in Wikidata with a ShEx schema E395 from a list" video 1) that use your csv file mentioned above 1) runs through the list and try to validate them with E395 1) As a lot of people you have in your list didnt appear in the books --> we have a source problem I also tried to compare your list with the book see MissingInBooks.csv

image image
salgo60 commented 1 year ago

Konstigheter

1) lmp t wd Q6009646 SPA

image
salgo60 commented 1 year ago

I am away for another longer kayak trip, guess back in 2-3 weeks.... assign this to me salgo60 if you need support

I feel we should

Wikidata sources on Swedish PM people using property "described by source" Property:P1343

image

image

image image
fredrik1984 commented 1 year ago

Yes, we have to decide what to call these MPs that are "partipolitiskt obunden", "opolitisk", etc. We will consult the Riksdag Library about this after the summer.

Thanks for helping out by adding parties to the MPs in the above list, this helps us a lot! And good luck on your kayak trip!

salgo60 commented 1 year ago

Yes, we have to decide what to call these MPs that are "partipolitiskt obunden", "opolitisk", etc. We will consult the Riksdag Library about this after the summer.

@fredrik1984 the problem I see right now is that we are waiting on something and we see odd things in the data as I mentioned above - we need to agree on an approach how to mark the data as questionable or need more care... --> when we get the way forward it will be easier to find those oddities and correct them

Focus list Välfärden analyserad - parti created

To get something started I mark those WD records were I see a potential problem with P5008 on focus list of Wikimedia project = Q120143028 - Välfärden analyserad - parti

imageimage

Feels like a good pattern

A pattern like this feels good to "mark" objects that are cared about in different research projects...

fredrik1984 commented 1 year ago

Hope you have had a nice summer with a lot of kayaking @salgo60! Just a question regarding this issue of adding party to the top known MPs but with unknown party: were you willing to add party to these MP wikidata posts?

salgo60 commented 1 year ago

@fredrik1984 this is the list with potential problems... plus all the records regarding communists also feels need some more care...

Focus list Välfärden analyserad - parti created

To get something started I mark those WD records were I see a potential problem with party P5008 on focus list of Wikimedia project = Q120143028 - Välfärden analyserad - parti

image
fredrik1984 commented 1 year ago

Thanks, I will try to take a look at this soon!

BobBorges commented 1 year ago

Closing as duplicate of #349 and #359