(cross-posting from wikitech-l)
Today we published an announcement on the Wikimedia blog marking the official launch of
revision scoring as a service
<https://meta.wikimedia.org/wiki/Research:Revision_scoring_as_a_service> and I
wanted to say a few words about this project:
Blog post:
https://blog.wikimedia.org/2015/11/30/artificial-intelligence-x-ray-specs/
<https://blog.wikimedia.org/2015/11/30/artificial-intelligence-x-ray-specs/>
Docs on Meta:
https://meta.wikimedia.org/wiki/ORES
<https://meta.wikimedia.org/wiki/ORES>
First off: what’s revision scoring
<https://meta.wikimedia.org/wiki/Research:Revision_scoring_as_a_service#Rationale>?
On the surface, it’s a set of open APIs allowing you to automatically “score” any edit and
measure their probability of being damaging or good-faith contributions. The real goal
behind this project, though, is to fix the damage indirectly caused by vandal-fighting
bots and tools on good-faith contributors and to bring back a collaborative dimension to
how we do quality control on Wikipedia. I invite you to read the whole blog post
<https://blog.wikimedia.org/2015/11/30/artificial-intelligence-x-ray-specs/> if you
want to know more about the motivations and expected outcome of this project.
I am thrilled this project is coming to fruition and I’d like to congratulate Aaron
Halfaker <https://wikimediafoundation.org/wiki/User:Ahalfaker> and all the project
contributors
<https://meta.wikimedia.org/wiki/Research:Revision_scoring_as_a_service#Team> on
hitting this big milestone: revision scoring started as Aaron’s side project well over a
year ago and it has been co-designed (as in – literally – conceived, implemented, tested,
improved and finally adopted) by a distributed team of volunteer developers, editors, and
researchers. We worked with volunteers in 14 different Wikipedia language editions and as
of today revision scores are integrated
<https://meta.wikimedia.org/wiki/Research:Revision_scoring_as_a_service#Tools_that_use_ORES>
in the workflow of several quality control interfaces, WikiProjects and 3rd party tools.
The project would not have seen the light without the technical support provided by the
TechOps team (Yuvi in particular) and seminal funding provided by the WMF IEG program and
Wikimedia Germany.
So, here you go: the next time someone tells you that LLAMAS GROW ON TREES
<https://en.wikipedia.org/w/index.php?diff=prev&oldid=642215410> you can
confidently tell them they should stop damaging
<http://ores.wmflabs.org/scores/enwiki/damaging/642215410/> Wikipedia.
Dario
Dario Taraborelli Head of Research, Wikimedia Foundation
wikimediafoundation.org <http://wikimediafoundation.org/> •
nitens.org
<http://nitens.org/> • @readermeter <http://twitter.com/readermeter>