Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Fwd: We need your help - SUMO localization experiment

1 view
Skip to first unread message

T.M. van den Broek

unread,
May 3, 2019, 6:22:14 AM5/3/19
to community-dutch


-------- Oorspronkelijke bericht --------
Onderwerp: We need your help - SUMO localization experiment
Datum: 2019-04-29 14:56
Afzender: Rubén Martín <nuke...@mozilla.com>
Ontvanger: tim_...@mozilla-nl.org, jetter...@gmail.com,
tpan...@live.nl



Hello everyone,
I wanted to get in touch with you all to talk about the Dutch SUMO
localization and an exciting experiment we want to run during the coming
months.
Summary:

*

We need your help between May 13rd and May 24th.
*

We will be experimenting with new localization technologies.
*

This can potentially engage new people and grow your communities.
*

This will help you cover the missing articles today on SUMO with an
improved experience.

It’s a known reality that the demand for new and updated localization
on SUMO has been increasing in the past years, Mozilla is bigger and we
are creating and testing new concepts and products faster than even (and
this will keep increasing in the future), which also translates into a
lot of new and updated documentation for these products that is as
important for users experience as the product itself.
We also value and care about your valuable volunteer time, you are
devoting your free time to help Mozilla and its product be better and
drive its mission in your language, and we want to ensure that
contributing to Mozilla is something fun and rewarding, that’s why we
want to optimize the experience.
Dutch is really important for Mozilla because we know there are a lot of
people speaking your language that actively use and benefits from the
products and we want to ensure all of them have the best experience.
We have the assumption that if we are able to improve the localization
experience we will save you a lot of your time, you will enjoy the
experience more and we will be able to do more with less. In order to
validate this assumption we want to run a test with you to see if
that’s true, we don’t want to make a decision yet, we want to test
with your help if we are walking in the right direction.
For that reason we have been talking with a lot of localization experts
and engaging with external organizations like Wikimedia with also a lot
of experience in this field, and it was clear that we can unlock a lot
of value by testing a Machine Translation approach that is controlled
and supervised by volunteers.
The current status of machine translation technologies has evolved a lot
in the past years and today we are able to produce super accurate
localization drafts that just take a few minutes for a human to
proofread and validate. Based on our external research and conversations
with other volunteer-based organization this can help you save 10 times
the effort you invest into getting an article ready in your language
today.
We also believe that this can help your communities grow, engaging
existing and new people who might be interested in a low effort work and
that might not need to be fluent in English to contribute.
Your collaboration in this experiment is fundamental, you will help
figure out if this approach is good or not and influence the SUMO
localization strategy in the coming years.
What would this look like?
Our plan is to create a list of articles that are currently outdated or
missing in your locale, run them through machine translation and import
them back for your review in the platform. This is very important, we
won’t publish anything that hasn’t been reviewed and approved by you
and I will send you a notification once these articles are ready for
review.
For this small experiment in Dutch we will be running it for Firefox
Desktop, Firefox Android and Firefox iOS, which have a huge market
importance in your language.
The time investment needed to help here would be around 3-5 minutes per
article, and our estimations are that we are talking about 20 max.
articles per locale. We would like to run the proofreading phase between
May 13rd and May 24th and we will be running a small survey after the
experiment to know your impressions.
Are you ready to help? Do you know someone who might be also interested
in your local community? Do you have any other questions? Please reply
to this email :-)
Thanks so much for your contributions!
Cheers.

--
Rubén Martín
Global Communities Strategist
https://mozillians.org/u/nukeador/

--
Mozilla Rep Mentor
Mozilla-nl.org
signature.asc

Fjoerfoks

unread,
May 11, 2019, 3:16:03 PM5/11/19
to Dutch Mozilla community
Hallo alle leden van deze mailinglijst,

Hierbij nogmaals een oproep om na te denken over het meewerken aan
onderstaand experiment.
Het idee is om nieuwe artikelen automatisch te laten vertalen, waarna er
nog een beoordeling plaats moet vinden.
Minder werk dan alles opnieuw te bedenken, maar dat hangt volledig af van
de geleverde kwaliteit.
Als je je geroepen voelt, laat dat dan even weten. Het experiment begin
waarschijnlijk deze week al.

Met vriendelijke groet,
Wim

Op vr 3 mei 2019 om 12:22 schreef T.M. van den Broek <
tim_...@mozilla-nl.org>:
> Mozilla-nl.org_______________________________________________
> community-dutch mailing list
> communi...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/community-dutch
>

Martijn Weisbeek

unread,
May 11, 2019, 4:44:52 PM5/11/19
to Dutch Mozilla community
Wim, Tim Maks,

Ik denk zeker dat we dit moeten uitproberen, om te zien of het bevalt.

Vooral de combinatie van Machine Translation and human proof-reading
spreekt mij aan:
- de machine voor het domme vertaalwerk
- en de menselijke factor om het helemaal goed te laten lopen

Daarmee blijft het eindresultaat toch voldoen aan de kwaliteit die wij
wensen en het zou ook tijd moeten besparen.
Daar kunnen we moeilijk op tegen zijn.

Persoonlijk ben ik niet direct een fan van geheel geautomatiseerde
vertalingen, want dat zie ik wel bij de KnowledgeBase-artikelen van
Microsoft die vaak geautomatiseerd in het Nederlands worden omgezet.
Natuurlijk scheelt ze dat tijd en zo hebben ze de teksten direct in alle
talen beschikbaar. Maar als ik dan naar die kwaliteit kijk, dan schakel ik
toch snel terug naar de Engelse tekst. Dan weet ik tenminste hoe ze het
bedoeld hebben.
Dezelfde soort automatische vertalingen zie ik ook bij de Nederlandse
beschrijvingen van apps in Google's Play Store (voor Android-apps). Ik neem
aan dat Microsoft een eigen vertaalmachine heeft en Google ook, maar beiden
produceren bij tijd en wijle dus belabberde teksten.

Als we dergelijke teksten zouden moeten redigeren, dan kunnen we misschien
net zo goed from-scratch de tekst blijven vertalen.
Maar mede daarom wil ik Mozilla hierin wel een kans geven. Laat ze maar
aangeven welke kwaliteit geautomatiseerde vertaling ze kunnen leveren.
Dan kunnen we zien wat we daaraan nog moeten doen en of dat tijd bespaart.

Ik denk dat we dit met een klein groepje vanuit Mozilla-NL moeten
beoordelen om te zien of dit ons helpt.
Ik ben doorgaans niet betrokken bij de vertalingen van teksten vanuit
Mozilla, maar wil hieraan mijn steentje bijdragen als dat dit project kan
helpen.

--
Met vriendelijke groet,

Martijn


Op za 11 mei 2019 om 21:16 schreef Fjoerfoks <fryske...@gmail.com>:
0 new messages