Re: [Wikimedia-l] Giving Commons a bigger public

2020-05-29 Thread Gerard Meijssen
Hoi,
I have been playing with the tool for some time and it is already great; it
works and it is a great basis to expand on. Expand its functionality and
data quality.

I added a label in Urdu in Wikidata for "Mujaddid Ahmad Ijaz" and once the
synchronisation was done, I found a wealth of pictures for him. That IS
powerful and demonstrates how labeling in Wikidata improves results in
Commons. I looked for "Frans Vera", a Dutch ecologists, not really finding
him in a lot of noisy results. There was one picture of him, I linked it to
his Wikidata item and after the synchronisation he was the first result.
That IS powerful because it shows how the linking to Wikidata of pictures
improves results.

As it is, it has important qualifications over Hay's tool and what Hay has
done is breaking ground. Now we know about the official proof of concept;
it has important advantages;
* it is internationalised and gets localised daily.
* it is already included in Commons so it will/must scale.
* it is supported by the WMF

One key question for me is; do we know its use. Do we know in what language
it is used? When people start using it for real, will we know? Commons has
enormous potential and it is now for us to make this a reality. One
challenge will be to convince me to use it. I do want to be convinced and I
am seeking for the arguments but in the end the proof of the pudding is in
the eating.

Keegan, I understand that as a technical guy you prefer to take it slowly,

For the WMF this is probably the biggest opportunity to remedy much of its
bias against all the "other" languages. I will blog and write about my
experiences because this has the potential to transform us into a truly
multi lingual movement as we now have ties that binds everything together.
Ties that are in and of themselves are useful in any language.
Thanks,
GerardM

On Fri, 29 May 2020 at 18:57, Keegan Peterzell 
wrote:

> On Sun, May 24, 2020 at 8:12 AM Michael Peel  wrote:
>
> >
> > It’s worth remembering that this functionality is built in to Commons,
> > it’s just not as user-friendly. From the example below, if you put
> > "haswbstatement:P180=Q191931” into the Commons searchbox, you will get
> the
> > same results. Thanks to the structured data on commons project+team!
> >
> >
> This is true.
>
> The Structured Data team is working on a media search prototype that is
> similar in function to Hay's tool. It's in the very earliest of early
> stages, that is to say that it works, and the team would like to hear
> feedback.
>
> Have a look over the project page if you're interested to see what a tool
> like Hay's could look like on Commons itself:
> https://commons.wikimedia.org/wiki/Commons:Structured_data/Media_search
>
> Comments welcome on the talk page, I'm slowly spreading the word about
> this.
>
> --
> Keegan Peterzell (he/him)
> Community Relations Specialist
> Wikimedia Foundation
> ___
> 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,
> 
___
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, 


Re: [Wikimedia-l] June 4 1800 Maggie Dennis office hour (with a twist)

2020-05-29 Thread Ciell Wikipedia
This will be very interesting, thank you Maggie.
(To check your local time for this office hour: click here
)


Vriendelijke groet,
Ciell


Op do 28 mei 2020 om 14:03 schreef Maggie Dennis :

> Hello, all.
>
> With the Board’s recent statement, this seems like a good time to launch
> the quarterly office hours I’ve been wanting to create for people who want
> to talk to me about issues involving “community resilience and
> sustainability,” including the work of Trust & Safety, whom I oversee.
> (after months of wanting to do this I’ve come to realize that I will always
> be "too busy" to feel like it's the perfect time for this. So I’m going to
> do it now anyway!)
>
> There’ve been requests to make office hours more personal, so I will host a
> Zoom hangout where people can join me, but I'll also take questions from
> Telegram and IRC.[1] I know that finding an hour that works for everybody
> is not going to happen, and I know from past office hours I’ve been
> involved in that I may get far more questions than I can answer (or,
> contrarily, nothing at all :)). Nevertheless, I will do my best to answer
> questions posed to me in that hour by Wikimedians in good standing (not
> Foundation or community banned) and to follow up in writing with any I
> don’t have time for over the next few days or week or so, time allowing. I
> might aggregate similar questions into a kind of FAQ. We’ll publish notes,
> anonymizing those who’ve asked questions, after.
>
> I do, however, have the following caveats:
>
>-
>
>I can’t and won’t discuss specific Trust & Safety cases. Instead, I can
>discuss Trust & Safety protocols and practices and approaches as well as
>some of the mistakes we’ve made, some of the things I’m proud of, and
> some
>of the things we’re hoping to do.
>-
>
>I will not respond to comments or questions that are disrespectful to
>me, to my colleagues, or to anyone in our communities. I can talk
> civilly
>about our work even if you disagree with me or I disagree with you. I
> won’t
>compromise on this.
>
>
> I’m not sure if I will stick with Zoom as the way I do office hours
> forever, but I am responding to some requests for spoken interaction while
> also trying to provide text options for those who prefer. I admit to being
> a little camera shy myself, so this is a challenge for me! If I embarrass
> myself too badly, I may retreat to the safety of text in future.
>
> I was hoping to have the Zoom link already, but while that’s being
> expedited by our office technology team, I don’t have it yet. I wanted to
> give interested people notice as soon as I knew the time. I’ll follow up
> with links again at least two hours in advance.
>
> The meeting will be on June 4th at 1800 UTC.
>
> I hope to see you there.
>
> Best,
>
> Maggie
>
> [1] Zoom link; Telegram link: https://t.me/joinchat/DOlGIB1FRLUWqW9iB3qfTQ
> ;
> directions for participating in IRC:
> https://meta.wikimedia.org/wiki/IRC_office_hours#How_to_participate
>
> --
> Maggie Dennis
> Vice President, Community Resilience & Sustainability
> Wikimedia Foundation, Inc.
> ___
> 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,
> 
___
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, 


Re: [Wikimedia-l] Giving Commons a bigger public

2020-05-29 Thread Keegan Peterzell
On Sun, May 24, 2020 at 8:12 AM Michael Peel  wrote:

>
> It’s worth remembering that this functionality is built in to Commons,
> it’s just not as user-friendly. From the example below, if you put
> "haswbstatement:P180=Q191931” into the Commons searchbox, you will get the
> same results. Thanks to the structured data on commons project+team!
>
>
This is true.

The Structured Data team is working on a media search prototype that is
similar in function to Hay's tool. It's in the very earliest of early
stages, that is to say that it works, and the team would like to hear
feedback.

Have a look over the project page if you're interested to see what a tool
like Hay's could look like on Commons itself:
https://commons.wikimedia.org/wiki/Commons:Structured_data/Media_search

Comments welcome on the talk page, I'm slowly spreading the word about this.

-- 
Keegan Peterzell (he/him)
Community Relations Specialist
Wikimedia Foundation
___
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, 


Re: [Wikimedia-l] An encyclopedia must be conservative (?)

2020-05-29 Thread Ziko van Dijk
Hello people, thanks for the reactions!

I actually did not mean conservative in a strict political sense, and I am
a big fan of Reagle's book. It seems to me
that some people in the movement identify strongly with the (political)
term "progressive", and, depending on their personal
circumstances, that can be very understandable. The risk is that their
thinking about WP, WM and knowledge is very much
predetermined by these political views.

On the other hand, I see the risk that some other members of the movement
think that they - and the movement - are
automatically "modern" because they use wikis. With this attitude they may
not see much need for change.

Allow me some comments between the - very interesting - lines. See below

Kind regards
Ziko


Am Fr., 29. Mai 2020 um 09:43 Uhr schrieb Amir E. Aharoni <
amir.ahar...@mail.huji.ac.il>:

> Aspect1: Fact-checking, trust, and reliability
> In non-wiki encyclopedias the writers
> are selected by the publisher: the publisher trusts the writers, and the
> readers trust the publisher's brand.  [...] The Wikpiedia attitude
> to sources, known as "Verifiability" in the English Wikipedia, solidified
> around 2005. It makes a lot of sense for a wiki encyclopedia, and it is one
> of our cornerstones, at least in the larger languages. (The details of the
> policy in each language may be different, but the general idea is the same.
> If it's significantly different in your language, please tell me.)
>

Indeed, cornerstone is a great word here. We don't check the contributors
as persons,
therefore we have to externalize the truth-question. So what does a
co-contributor check when seeing
the edit of a contributor? Not whether the contributor has thematic or
scientific competences but only whether she
has "publishing" or wirting competences - meaning, whether she is capable
of selecting good literature, using it
and writing accordingly an unpersonal encyclopedic text.

Within the wiki, it would be difficult to check whether she has scientific
competences. But it is possible to
check within the wiki whether she has writing competences, because we see
these writing competences in her previous
edits. So in Wikipedia, it works that we only care for the "wiki identity"
and "wiki status" of a contributor.

By the way, it would be an interesting research question: are there
Wikipedias that significantly differ from this cornerstone
(and other cornerstones)? In my own comparison between WP in EN, DE, NL, AF
and FY I did not find such a difference.


no access to academic publishing? Some people propose relaxing the demand
> for external reliable source for such topics, and while I'm totally on
> board with the social justice aspect of this attitude, it doesn't suggest a
> solution to the trust problem: some people will use it to enrich Wikipedia
> with information that can't be found elsewhere, but some people may abuse
> it to add made up stuff.
>

Yes, indeed. My thought was: if we allow sources of probably lesser
quality, e.g.
"grey literature" for marginalized people as an article topic, what would
that mean? That we
find it okay that a Wikipedia article about a woman is less reliable than
an article about a man?



> I have a proposed solution for this problem, and although some people would
> disagree, I call it conservative: Keep the demand for verifiability, and
> help people who have been historically disadvantaged get access to trusted
> academic institutions and conduct and publish their research outside of
> Wikipedia first.
>

Interesting. Some might ask whether this is a task of the Wikimedia
movement. (It all would depend on suitable partners and what the
role of the movement would be.)

If we see there a tunnel like this: reality - primary sources - secondary
sources - tertiary sources, then the
solution would not be at the final stage (tertiary source = Wikipedia and
its rules), but earlier, at the stage secondary sources
where there is a social filter.



> Aspect 2: Technology
> reasonably modern design principles and implementations. We are outdated in
> some ways: [...] We shouldn't be *too* progressive, though [...]
>
Talk pages are a particularly curious kind of disaster. Many Wikipedians
> tend to be very conservative about them and don't want any technology
> changes in them, but talk pages are not a continuation of any previous
> tradition of encyclopedic writing or of Internet culture—they are
> Wikipedia's own invention.


Good point, Amir. This is a dimension I have not looked at very thoroughly
in my research.
It blends in with the general question how software is interconnected to
content and to
the behaviour of the contributors. Maybe the current state of the software
used has for
some Wikipedians the function of a cultural marker or element of
self-identity.


> Aspect 3: Presentation style

[...]

> Like the bold font, it is also a typographic tradition. It
> has gotten out of hand in Wikipedia thanks to otherwise good things like
> the 

Re: [Wikimedia-l] An encyclopedia must be conservative (?)

2020-05-29 Thread Amir E. Aharoni
Most people in the world (or at least in the U.S.) use the terms
"conservative" and "progressive" when talking about politics, and associate
them with bundles of viewpoints on society, economics, religion, and so on.
The political aspect is partly relevant to Wikipedia, too, but if we just
take them as words with literal meanings, we'll have to talk about some
other aspects, too. Here are the ones I can think of, in a mostly-random
order:

Aspect1: Fact-checking, trust, and reliability

Fact-checking, trust, and reliability on Wikpiedia should be conservative,
but in a way that is thoughtful and open to challenging itself. It's a
difficult and often overlooked point. In non-wiki encyclopedias the writers
are selected by the publisher: the publisher trusts the writers, and the
readers trust the publisher's brand. I'm intentionally not saying "printed"
or "old" encyclopedias, but "non-wiki" encyclopedias. They are still being
produced, in print and digitally—see my Wikimania 2014 talk[1] for just one
example.

Our wiki model wants to let everyone write, and writers are not pre-vetted,
so our solution for trust is demanding reliable sources, which is why
Wikipedia articles in many languages have a lot of footnotes. Other
encyclopedias usually don't have footnotes, although some do have a
"further reading" or "bibliography" at the end of some articles, but they
are provided for further research and not for proof. The Wikpiedia attitude
to sources, known as "Verifiability" in the English Wikipedia, solidified
around 2005. It makes a lot of sense for a wiki encyclopedia, and it is one
of our cornerstones, at least in the larger languages. (The details of the
policy in each language may be different, but the general idea is the same.
If it's significantly different in your language, please tell me.)

The problem with this attitude is that it outsources trust to other
publishers: non-wiki encyclopedias, academic journals, newspapers and news
sites, and occasionally other sources. The better-known issue with it is
deciding which external sources are reliable. The less-known, but perhaps
even trickier issue is what to do about topics that should be covered in an
encyclopedia, but about which there is no coverage in what Wikipedia
editors would call "reliable sources" because of systemic bias, that is
because the people who are involved with the topic had historically less or
no access to academic publishing? Some people propose relaxing the demand
for external reliable source for such topics, and while I'm totally on
board with the social justice aspect of this attitude, it doesn't suggest a
solution to the trust problem: some people will use it to enrich Wikipedia
with information that can't be found elsewhere, but some people may abuse
it to add made up stuff.

I have a proposed solution for this problem, and although some people would
disagree, I call it conservative: Keep the demand for verifiability, and
help people who have been historically disadvantaged get access to trusted
academic institutions and conduct and publish their research outside of
Wikipedia first. The WMF and its partners can do it. It's not easy, but I
just don't see any other solution to the trust issue. I call this attitude
"conservative" because I want to preserve the trust in external knowledge
institutions, and keep the "outsourcing". It's not exactly what the current
strategy recommendation[2] says, and I respectfully doubt that that
recommendation is going to work.

Aspect 2: Technology

Should be reasonably progressive, of course, in the sense of using
reasonably modern design principles and implementations. We are outdated in
some ways: talk pages are a disaster, the jQuery JavaScript framework is
quite old (and is being gradually replaced), many templates are too
difficult to maintain, code review and feature deployment are not as robust
as they should be, and there are many other issues.

We shouldn't be *too* progressive, though: we should not jump on every
buzzword bandwagon and not to change design concepts and development
frameworks every year, as some sites do. It's probably good that we are not
jumping on the blockchain bandwagon at all, and that we are jumping on the
artificial "intelligence" bandwagon in a careful and measured way (ORES is
helpful, but keeps the human decision in the loop).

Talk pages are a particularly curious kind of disaster. Many Wikipedians
tend to be very conservative about them and don't want any technology
changes in them, but talk pages are not a continuation of any previous
tradition of encyclopedic writing or of Internet culture—they are
Wikipedia's own invention. Is it good that the community, or at least some
parts of it, is so conservative about it? Not really, and it causes serious
damage as time goes by, but arguing with these passionate people is
challenging.

Aspect 3: Presentation style

Should be conservative in the sense of continuing a centuries-old tradition
of writing