Has anyone tried to use a Siamese neural network for image classification
at Commons? I don't know if it will be good enough to run in autonomous
mode, but it will probably be a huge help for those that do manual
Imagine a network providing a list of possible categories, and the user
just ticks off usable categories.
A Siamese network can be learned by using a triplet loss function, where
the anchor and the positive candidate comes from the same category, and the
negative candidate comes from an other category but are otherwise close to
Output from the network is like a fingerprint, and those fingerprints can
be compared to other images with known fingerprints, or against a
generalized fingerprint for a category.
John Erling Blad
I have started an RfC here: https://meta.wikimedia.org/
This RfC is not intended to cast a negative light on communities and
individuals who are making good-faith efforts to welcome new users by
placing welcome messages on the new users' talk pages. The emphasis on the
RfC is on whether changes should be made to the privacy of account creation
logs, which appears to affect the privacy of logged-in users who are
reading Wikimedia wikis which they have not previously visited while logged
in. I made proposals in the RfC regarding how new users may be welcomed.
Please find here  a link to our draft Progress Report. This report will
be further reviewed for language, metrics, and other visualisation before
submission. We would like to thank you for the support during the drafting
of this report.
Do let us know if there is anything else,
Thanks and Regards,
*ANANTH SUBRAY P V(ಅನಂತ್)*
Access to Knowledge program
The Centre for Internet & Society
As many of you know, in October we concluded phase 1 of the movement
strategy process. The result was a final draft of the strategic
direction, summarizing the hundreds of conversations that took place all
over the world, on wiki and off, about where we as a movement want
Wikimedia to go next. Many communities and individuals have signed on to
the direction, expressing their support for the guide we collectively
created for our future.
The Board recognizes the challenges and opportunities that lie ahead in
making this direction a reality. In the next phase of the movement
strategy, we will get into more of the specifics of how to make that
happen. With that in mind, we would like to share a statement setting forth
our commitment to the future of Wikimedia, and a clear mandate for the
Wikimedia Foundation to invest the resources necessary to support the
growth and evolution required for the next chapter of Wikimedia’s future.
Our statement is included below, and on Meta-Wiki, where it has been set up
The Board greatly appreciates all of the time and energy that thousands of
people have put into the movement strategy process. Special appreciation
goes out to the members of the community who stepped up to help lead local,
language, or global organizing efforts. We are not done yet, but what we
have created is something that we should all be proud of -- for the process
of how we got to this direction, as much as for the direction itself.
On behalf of the Wikimedia Foundation Board of Trustees,
Christophe Henner, Board Chair
At our most recent Board meeting on the 18th of November, the Wikimedia
Foundation Board focused much of our discussion on the needs and goals of
the Wikimedia 2030 movement strategy process. We carefully considered
the next steps we as a movement, and the Wikimedia Foundation in
particular, need to take to build for our future.
The Board is committed to ensuring the vision outlined in the Wikimedia
2030 process becomes a reality. To support this direction and the future of
the Wikimedia movement, it is our belief that the Wikimedia Foundation must
expand its resources through healthy, sustainable practices. To this end,
we want to give a clear mandate for the Wikimedia Foundation to invest the
resources necessary to support the growth and evolution required for the
next chapter of Wikimedia’s future.
We specifically recommend that the Foundation:
Increase investment in Foundation staffing and other means of support
for the movement direction, sufficiently resourcing product, technology,
and community health commitments in particular;
Support and engage with individuals, groups, and organizations,
especially within the Wikimedia movement, to further develop their
capacities, including the specific needs of emerging communities;
Support the fundraising team in raising additional funds beyond what is
called for in the annual plan to prepare us for future growth;
Increase revenue as needed to support investment and growth;
Explore alternative revenue streams for the Foundation and movement; and
Undertake any capacity expansion in a healthy and sustainable way that
anticipates current and future needs.
Based on anticipated need and past performance, we envision an annual
budgetary growth rate of 10–20% over the next several years.
The Board takes seriously its responsibility to the Foundation and by
extension, to the global Wikimedia movement. We believe this mandate will
better ensure we can realize the future we all have outlined as part of
Wikimedia 2030. We are committed to ensuring the sustainable growth and
success of our movement, and the Wikimedia Foundation’s role in supporting
Chair of the board of trustees
twitter *@schiste* skype *christophe_henner*
Thanks, Jonathan. Do you think we can convince Katherine Maher to
agree to enforce the Creative Commons Attribution requirements? There
is no doubt it would aid both editor recruitment, and as you point
out, morale too.
I also want to ask her about:
(2) survey metrics:
(3) benchmarking investment performance against institutional
endowment-grade mutual funds and studies of endowment performance:
and e.g., https://institutional.vanguard.com/iam/pdf/EndowmentPerformanceResearch.pdf
Related: endowment size required for full sustainability;
(4) testing replacing the pencil icon with the word "" on mobile;
(5) intelligibility remediation on Wiktionary as a Foundation
technology development project;
(6) systemic review of bias in economics articles; and
(7) an ongoing top performers' invitational essay contest for the
Katherine, what are your opinions on those recommendations?
Can (6) and (7) be combined?
On Sun, Jan 28, 2018 at 10:50 PM, Jonathan Cardy
> I wouldn't express it quite so bluntly, but agreed at a time when editing seems to have stabilised again after the 2015/16 rally, shifting the Foundation to a strategy of promoting compliance with both BY and SA would address a lot of problems. It is probably demotivating for editors to see their work used without attribution, and whilst a link back to Wikipedia is not as going to be as good as an edit button, we are greatly limiting ourselves if we rely on people coming directly to our sites and treat every extract from our sites as CC0 or Fair Use.
> A few legal letters and maybe a court case a year should be easily affordable for the WMF and an excellent investment.
>> Message: 2
>> Date: Sun, 28 Jan 2018 00:12:43 +0000
>> From: James Salsman <jsalsman(a)gmail.com>
>> To: Wikimedia Mailing List <wikimedia-l(a)lists.wikimedia.org>
>> Subject: Re: [Wikimedia-l] Copyright enforcement?
>> Content-Type: text/plain; charset="UTF-8"
>> Attribution is often considered impractical, but providing the source
>> date along with e.g. the article name can be used to derive the
>> attribution, so it should be required. It's not just a good idea to
>> require this information from content re-users like Amazon, Apple, and
>> Google, but doing so will help encourage those who find issues to
>> If the Foundation doesn't make attribution or at least article date a
>> requirement, then they are actively opposing editor recruitment.
>>> On Sat, Jan 27, 2018 at 7:34 PM, The Cunctator <cunctator(a)gmail.com> wrote:
>>> The copyright requirement isn't attribution; it's attribution and copyleft
>>> retention for derived works.
> 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(a)lists.wikimedia.org
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:firstname.lastname@example.org?subject=unsubscribe>
I wouldn't express it quite so bluntly, but agreed at a time when editing seems to have stabilised again after the 2015/16 rally, shifting the Foundation to a strategy of promoting compliance with both BY and SA would address a lot of problems. It is probably demotivating for editors to see their work used without attribution, and whilst a link back to Wikipedia is not as going to be as good as an edit button, we are greatly limiting ourselves if we rely on people coming directly to our sites and treat every extract from our sites as CC0 or Fair Use.
A few legal letters and maybe a court case a year should be easily affordable for the WMF and an excellent investment.
> Message: 2
> Date: Sun, 28 Jan 2018 00:12:43 +0000
> From: James Salsman <jsalsman(a)gmail.com>
> To: Wikimedia Mailing List <wikimedia-l(a)lists.wikimedia.org>
> Subject: Re: [Wikimedia-l] Copyright enforcement?
> Content-Type: text/plain; charset="UTF-8"
> Attribution is often considered impractical, but providing the source
> date along with e.g. the article name can be used to derive the
> attribution, so it should be required. It's not just a good idea to
> require this information from content re-users like Amazon, Apple, and
> Google, but doing so will help encourage those who find issues to
> If the Foundation doesn't make attribution or at least article date a
> requirement, then they are actively opposing editor recruitment.
>> On Sat, Jan 27, 2018 at 7:34 PM, The Cunctator <cunctator(a)gmail.com> wrote:
>> The copyright requirement isn't attribution; it's attribution and copyleft
>> retention for derived works.
Users on other projects are complaining about the welcome messages at
arwiki. A bot at that project are welcoming people that has no activity at
that project at all. The bot operator claims the activity is valid, but I
can't see that this is a well-behaving bot at all.
I suspect the bot is welcoming every user it can find, but using user
accounts from central login and not users that has local contributions at
Can someone shut down the bot until the user fix the spam problem.
Each year the Wikipedia Library team, in collaboration with GLAM folks
(Galleries, Libraries, Archives & Museums) all around the world, asks a
simple favor: Give Wikipedia the gift of a citation for its birthday.
This year Wikipedia turns 17 and it's our third annual #1Lib1Ref campaign.
The 20-day campaign is now in its second week... What's happened so far?
Across the languages we are tracking, editors have added 2097 citations in
5 languages have made over 100 edits: English, French, Hebrew, Serbian and
View contributions by language:
On social media #1Lib1Ref is a hype-machine. 37 countries have made 2260
Twitter posts reaching 4.4 million potential readers 8.1 million times.
We invited librarians (and library-lovers) around the world to participate. Our
Wikimedia Blog post, written by #1Lib1Ref originator and GLAM Strategist
Alex Stinson, set the tone and made the case:
The blog included a visual 'explainer' video created by Wikipedia Library
team member Felix Nartey, narrated by famous U.S. Librarian Jessamyn West.
We promoted the Story of #1lib1ref on social media using an 8-part
Also we redesigned a simple new graphic and logo for the campaign:
Each year #1Lib1Ref grows and brings in more people from libraries, and
more people who believe that Wikipedia and librarians share a mission which
benefits from collaboration. It's an exciting part of our work and we're
very thankful to everyone who adds to its impact and success!
Join in. Make an edit. Teach someone how to add a citation. Share something
nice about the campaign this week in your social media circles. Help share
free knowledge. #1Lib1Ref. Pass it on...
Head of The Wikipedia Library
We are seeing a steady decrease of page views to our projects
(Wikipedia). Nov-Dec-Jan it is decreasing in a rate of 5-10%
(year-year), and for big languages like Japanese, Spanish close to 10%,
or some months even more 
Is there any insights of why this is so? Could it be that Google take
over accesses with their ever better way of showing results direct (but
then also with showing extracts of Wikipedia articles) . Or that our
interface on mobiles is inferior so we miss accesses from mobiles (now
being 54% of total). Or horror of horror that users look for facts on
all new sites with fake news instead of Wikipedia?