Hi Ariel, Denny, and Wikimedians,
I sent this at noon today, but it didn't go through to this email address, so am re-sending now from here.
Suggestions about how these email lists might work differently?
Best, Scott
Scott MacLeod worlduniversityandschool@gmail.com 12:01 PM (5 hours ago) to Wikimedia Thanks, Ariel, Denny, and Wikimedians,
Re your observations, I wonder, conceptually, if the end-to-end Translation approach of Google Translate would render what you suggest, Ariel (e.g. someone could translate an article from a language Wikipedia and this would produce a single translation i.e. "one article for all audiences"), whereas developing, conceptually, a different translator approach - i.e. from Language A to Language B - conceptually, and perhaps building such a hypothetical translator from Wikidata's Lexicographical project, could yield "information depending on their role in the communities of the speakers of a given language" (leaving aside the Wikimedia's ContentTranslation project at this stage).
And I wonder re 1), Denny and Ariel, beyond cultural contexts / differences / borders, and your Portugal and Brazil example, whether one might add in a country approach as well. Re 2 & 3) I wonder what role interpretation plays in writing a NPOV article in any given language, and vis-a-vis Abstract Wikipedia too, and how these NPOV interpretations offer benefits in their great diversity (in terms of Wikipedia's goal: "to build a world in which every single human being can freely share in the sum of all knowledge"), and the benefits of "displaying the local articles whenever available" for inter-lingual knowledge-generation.
Thanks for the very thoughtful questions and responses.
Cheers, Scott
On Mon, Dec 10, 2018 at 5:25 PM Denny Vrandečić vrandecic@gmail.com wrote:
Hi Ariel,
thanks for the very thoughtful question. I got asked this question every time I present it, and during the Blue Sky presentation this question - or a variation of it - was asked three times. It really is on top of people's mind!
My answer is half inconsistent, I am afraid, because I have by now come up with three ways to answer this question, and they contradict each other. So I am glad to hear more thoughts on it.
Here are the three answers:
- I think that language is a pretty bad delimiter to keep points of view
apart. Yes, sure, it allows the Japanese Wikipedia to offer a different description of World War 2 than the Korean Wikipedia has, but I am not sure that is entirely a good thing. We don't have two Wikipedias for Portugal and Brazil, they have to agree and what they say, but we have Wikipedias in Croatian, Serbian, Bosnian, Serbocroatian... and I am not sure that the outcome of this decision is fully positive. So, my main point is, if we really want to capture cultural differences, let's align the borders of the editions of Wikipedia along these cultures. But aligning the cultural borders solely along language borders is badly imperfect.
- But in general, I think that accepting that different Wikipedias should
have different contents are incompatible with our NPOV policy. Now we could have a lengthy discussion whether NPOV is a good policy or not. But in general, I'd really prefer to have all points of views being presented with their due weight in all languages, instead of using languages to represent a point of view only in one language, and have a different point of view in another language. I would love to be able to read both the Japanese and the Korean point of view on contentious issues between these two countries - as I can in Serbian and Croatian, because I can read both languages just fine
- but I think it is rather problematic that language barriers dictate the
point of view I have access to. In fact, in many cases, we can see in the English speaking Wikipedia how the very same editors from the say Croatian and Serbian Wikipedia come to a more balanced result in the English Wikipedia, which they wouldn't accept in their 'home' Wikipedias. Funny, isn't it?
- More importantly, and entirely disagreeing with #1 and #2, is that the
Abstract Wikipedia never suggests to replace the current language editions, but to fill up the gaps in any given language edition. So, if the Croatian Wikipedia really wants to go into details on Croatian folk songs and Croatian food items, they should be totally able to do so without having to feel bad that they might be missing basic information about South American countries and Australian Aboriginal cultures. In fact, what I hope is that each language edition can choose to display the renderings from the Abstract Wikipedia for most articles, and then they can concentrate on creating in-depth articles on the topics they really care about - local cities, cultures, traditions. I remember in the beginning when working on the Croatian Wikipedia - it feels weird to work on the article about a local dish if you're still missing articles on all chemical elements. How can I write an article about the town my mom lives in (pop. 148) if there is no article yet about the country of Mexico? The abstract Wikipedia has the ability to lessen that pressure and allow the local communities to focus on their interests more.
I assume that the solution that combines the three answers is that we will display the local articles whenever available, but be able to take a look at the article rendered from the abstract version (for contrast and comparison and maintenance). And if there is no local article, we would treat the article rendered from the abstract version like a proper article.
So, as said, the actual answer to your question is still work in progress, but I wanted to write down a first sketch towards the answer. I am also very happy to hear other people thoughts on this question. But all in all I think that going toward an Abstract Wikipedia will improve along all the dimensions discussed. But I am sure I missed important dimensions on this question.
Cheers, Denny
On Thu, Nov 15, 2018 at 11:15 PM Ariel Glenn WMF ariel@wikimedia.org wrote:
I want to add a caution about the idea of translating one article for all audiences. Even articles on some plants or animals will contain different information depending on their role in the communities of the speakers
of a
given language; how much more will articles about some politician or a religious custom vary depending on the presumed cultural context of the community of readers? Even sources vary according to the language of the project, with sources in the project language preferred for ease of verifiability. One of the strengths of multi-language Wikipedia is this very concept of a topic being presented in a fashion that is suitable to different communities of readers, and the language of the text is only
one
part of that.
On Fri, Nov 16, 2018 at 8:40 AM Leila Zia leila@wikimedia.org wrote:
Denny, thanks for writing and rewriting this piece. I finally got a
chance
to go through it end-to-end. Challenge accepted! :)
Here are a few early thoughts, and I look forward to discussing it with
you
and others further.
- I tend to agree with you that the challenges of artificial
intelligence
are a superset of the challenges of bringing to life the abstract Wikipedia. Quite a few items you list in "Unique advantages" section
make
the abstract-Wikipedia space more easily approachable.
- I agree with you that if we are to take the content of Wikipedia to
many
of the languages spoken in the world today, and engage their speakers
to
share in, the current model won't work/scale (at least soon enough).
- You've raised a great point about "Graceful degradation". A very nice
challenge.
- In "Unique advantages" you talk about "a single genre of text,
encyclopedias" and I wonder what it takes to expand our thinking to
include
images as well. Will we need to rethink your current construct?
Including
images is attractive for at least two reasons: Because in terms of
learning
people have different needs and we will likely need to (continue to) include images as we create the abstractions, but also because one can potentially think of images as representations that are already
abstract.
Best, Leila
-- Leila Zia Senior Research Scientist, Lead Wikimedia Foundation
On Tue, Nov 13, 2018 at 10:13 AM Dariusz Jemielniak <darekj@alk.edu.pl
wrote:
an interesting concept indeed!
dj
On Tue, Nov 13, 2018 at 5:36 PM Denny Vrandečić <vrandecic@gmail.com mailto:vrandecic@gmail.com> wrote: The extended whitepaper that was presented at the DL workshop is now available here:
http://simia.net/download/abstractwikipedia_whitepaper.pdf
Still not a proper scientific paper (no references, notv situated in related work), but going into a bit more detail on the ideas on the
first
paper published previously.
On Sat, Sep 29, 2018, 11:32 Denny Vrandečić <vrandecic@gmail.com
<mailto:
vrandecic@gmail.com> wrote:
Semantic Web languages allow to express ontologies and knowledge
bases
in
a way meant to be particularly amenable to the Web. Ontologies
formalize
the shared understanding of a domain. But the most expressive and widespread languages that we know of are human natural languages,
and
the
largest knowledge base we have is the wealth of text written in
human
languages.
We looks for a path to bridge the gap between knowledge
representation
languages such as OWL and human natural languages such as English.
We
propose a project to simultaneously expose that gap, allow to
collaborate
on closing it, make progress widely visible, and is highly
attractive
and
valuable in its own right: a Wikipedia written in an abstract
language
to
be rendered into any natural language on request. This would make
current
Wikipedia editors about 100x more productive, and increase the
content
of
Wikipedia by 10x. For billions of users this will unlock knowledge
they
currently do not have access to.
My first talk on this topic will be on October 10, 2018,
16:45-17:00,
at
the Asilomar in Monterey, CA during the Blue Sky track of ISWC. My
second,
longer talk on the topic will be at the DL workshop in Tempe, AZ,
October
27-29. Comments are very welcome as I prepare the slides and the
talk.
Link to the paper: http://simia.net/download/abstractwikipedia.pdf
Cheers, Denny
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.orgmailto: Wikimedia-l@lists.wikimedia.org Unsubscribe:
https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
<mailto:wikimedia-l-request@lists.wikimedia.orgmailto: wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>
-- ________________________________________________________ [http://crow.kozminski.edu.pl/minds.jpg]<
http://nerds.kozminski.edu.pl/%3E
prof. dr hab. Dariusz Jemielniakkierownik katedry MINDS (Management in Networked and Digital
Societies)
Akademia Leona Koźmińskiego http://NeRDS.kozminski.edu.pl http://nerds.kozminski.edu.pl/
Ostatnie artykuły:
- Dariusz Jemielniak, Maciej Wilamowski (2017) Cultural
Diversity
of
Quality of Information on Wikipedias< http://crow.kozminski.edu.pl/papers/cultures%20of%20wikipedias.pdf%3E Journal of the Association for Information Science and Technology 68:
2460–2470.
- Dariusz Jemielniak (2016) Wikimedia Movement Governance: The
Limits
of A-Hierarchical Organization< http://www.crow.kozminski.edu.pl/papers/wikimedia_governance.pdf%3E
Journal
of Organizational Change Management 29: 3. 361-378.
- Dariusz Jemielniak, Eduard Aibar (2016) Bridging the Gap
Between
Wikipedia and Academia< http://www.crow.kozminski.edu.pl/papers/bridging.pdf%3E Journal of the Association for Information Science and Technology 67: 7.
1773-1776.
- Dariusz Jemielniak (2016) Breaking the Glass Ceiling on
Wikipedia<
http://www.crow.kozminski.edu.pl/papers/glass-ceiling.pdf%3E Feminist Review 113: 1. 103-108.
- Tadeusz Chełkowski, Peter Gloor, Dariusz Jemielniak (2016)
Inequalities in Open Source Software Development: Analysis of
Contributor’s
Commits in Apache Software Foundation Projects<
http://journals.plos.org/plosone/article/asset?id=10.1371%2Fjournal.pone.015...
, PLoS ONE 11: 4. e0152976. _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.org Unsubscribe:
https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe