I sent a long email to Wikimedia-l and also made the same post to Meta. I published a new paper recently with a proposal for a multilingual Wikipedia and more, and, unsurprisingly, Wikidata plays a central role in that proposal. I am trying to have the discussion not to be too fragmented, so I hope it will happen on Meta or on Wikimedia-l, but I also wanted to give a ping here.
Stay safe! Denny
[1] https://lists.wikimedia.org/pipermail/wikimedia-l/2020-April/094621.html [2] https://meta.wikimedia.org/wiki/Wikimedia_Forum#Proposal_towards_a_multiling... [3] https://arxiv.org/abs/2004.04733
Thanks Denny! I'll read the PDF on arxiv tomorrow over my morning coffee(s).
Stay tuned.
Hey Denny,
I've finally managed to reach the conclusion of your paper. It's truly a lot to digest, especially for people like me who do not know much of these things, but it was a really thorough and interesting read.
One thing that got me thinking is the part about biases - please note that it is not a critique, merely a *very* confused, and possibly fairly terrible, thought about it :)
So, biases exist, they inform most of the discussion about delicate matters, and we're all "POV healthy carriers". In a way, this is fine to a certain limit. Many people, though, "refuse" (for a lack of better words) to acknowledge that in relation to themselves, and this is a huge problem to overcome when we need to find a way to establish a consensus about an as-much-as-NPOV-possible text about $subject.
My fear is that all our attempts, from Wikidata to your proposal, are extremely good at noticing where bias is or might be, where we should point our attention to, but still aren't enough to fight back the refusal to acknowledge a bias. We still need a way to find people willing to tackle this, or at least giving them enough motivation to solve this in our current working framework.
On the bright side, I can see so many applications of your project that I can't wait for it to happen. :)
L.
Il giorno mar 14 apr 2020 alle ore 03:10 Denny Vrandečić vrandecic@gmail.com ha scritto:
I sent a long email to Wikimedia-l and also made the same post to Meta. I published a new paper recently with a proposal for a multilingual Wikipedia and more, and, unsurprisingly, Wikidata plays a central role in that proposal. I am trying to have the discussion not to be too fragmented, so I hope it will happen on Meta or on Wikimedia-l, but I also wanted to give a ping here.
Stay safe! Denny
[1] https://lists.wikimedia.org/pipermail/wikimedia-l/2020-April/094621.html [2] https://meta.wikimedia.org/wiki/Wikimedia_Forum#Proposal_towards_a_multiling... [3] https://arxiv.org/abs/2004.04733
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
-- Luca "Sannita" Martinelli http://it.wikipedia.org/wiki/Utente:Sannita
Hey Luca,
thank you so much for taking the time to read the proposal in detail, and managing to get through it.
Yes, I agree, you are right. This proposal, just as Wikidata, is good at exposing and identifying biases, but not that good at actually fixing them. I also think that this is OK - the fixing should probably not be something a tool should do, but that is up to us as the communities to get it done.
In fact, as you can find in the Wikipedia@20 essay
https://wikipedia20.pubpub.org/pub/vyf7ksah
one of the considerations for Abstract Wikipedia is indeed to make it more explicit to see which biases are intentional in a language editions, and which ones are not, in the hope that we can then tackle certain biases in some language editions in a more targeted way.
Thank you for your kind words, and I am also very excited to get this thing moving! :)
Stay safe, Denny
On Fri, Apr 17, 2020 at 7:56 AM Luca Martinelli martinelliluca@gmail.com wrote:
Hey Denny,
I've finally managed to reach the conclusion of your paper. It's truly a lot to digest, especially for people like me who do not know much of these things, but it was a really thorough and interesting read.
One thing that got me thinking is the part about biases - please note that it is not a critique, merely a *very* confused, and possibly fairly terrible, thought about it :)
So, biases exist, they inform most of the discussion about delicate matters, and we're all "POV healthy carriers". In a way, this is fine to a certain limit. Many people, though, "refuse" (for a lack of better words) to acknowledge that in relation to themselves, and this is a huge problem to overcome when we need to find a way to establish a consensus about an as-much-as-NPOV-possible text about $subject.
My fear is that all our attempts, from Wikidata to your proposal, are extremely good at noticing where bias is or might be, where we should point our attention to, but still aren't enough to fight back the refusal to acknowledge a bias. We still need a way to find people willing to tackle this, or at least giving them enough motivation to solve this in our current working framework.
On the bright side, I can see so many applications of your project that I can't wait for it to happen. :)
L.
Il giorno mar 14 apr 2020 alle ore 03:10 Denny Vrandečić vrandecic@gmail.com ha scritto:
I sent a long email to Wikimedia-l and also made the same post to Meta.
I published a new paper recently with a proposal for a multilingual Wikipedia and more, and, unsurprisingly, Wikidata plays a central role in that proposal. I am trying to have the discussion not to be too fragmented, so I hope it will happen on Meta or on Wikimedia-l, but I also wanted to give a ping here.
Stay safe! Denny
[1]
https://lists.wikimedia.org/pipermail/wikimedia-l/2020-April/094621.html
[2]
https://meta.wikimedia.org/wiki/Wikimedia_Forum#Proposal_towards_a_multiling...
[3] https://arxiv.org/abs/2004.04733
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
-- Luca "Sannita" Martinelli http://it.wikipedia.org/wiki/Utente:Sannita
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata