Hey all,
I am thrilled to announce a new formal collaboration [1] between the
Wikimedia Foundation and a team of researchers at Stanford University,
aiming to design and improve strategies to detect potential sockpuppets in
Wikimedia projects.
<https://meta.wikimedia.org/wiki/Research:Sockpuppet_detection_in_Wikimedia_…>,
part of the Wikimedia Foundation's community health program [2].
Sockpuppetry is the use of more than one account on a social platform. It
is a major problem in Wikipedia, as it is frequently used for vandalism,
paid editing, pushing one's point of view into articles, and bypassing the
community guidelines. On English Wikipedia specifically, benign and
malicious uses are well defined [3]. Recent research has elicited deceptive
and non-deceptive behavior in online discussions, and identified that
deceptive sockpuppets primarily aim to create illusion of consensus [4].
Our aim in this project is to design machine learning algorithms to
identify potential sockpuppet accounts on English Wikipedia. We will
leverage data from previously identified sockpuppets [5] to train our
models for detection and create high precision models to identify new
sockpuppets. The expected outcome of this project is a set of open
algorithmic methods, and a report on their performance and limitations,
that could be integrated later on into tools to support community efforts
to identify and flag sockpuppet accounts.
I am excited to launch this collaboration with Srijan Kumar, Tilen Marc,
Jure Leskovec and their group at Stanford, who have a solid record of
research on this topic (and most recently also studied hoaxes in Wikipedia
[6]).
You can follow the progress on this project on its page on Meta
<https://meta.wikimedia.org/wiki/Research:Sockpuppet_detection_in_Wikimedia_…>,
where we'll be reporting the results, or chime in on the talk page
<https://meta.wikimedia.org/wiki/Research_talk:Sockpuppet_detection_in_Wikim…>
.
Dario
[1] https://www.mediawiki.org/wiki/Wikimedia_Research#Collaborations
[2]
https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Annual_Plan/2017-2018/…
[3] Wikipedia:Sock puppetry. https://en.wikipedia.org/wiki/
Wikipedia:Sock_puppetry
[4] An Army of Me: Sockpuppets in Online Discussion Communities. S. Kumar,
J. Cheng, J. Leskovec, V.S. Subrahmanian. *Proceedings of World Wide Web
conference*, 2017.
[5] Category:Wikipedia sockpuppets. https://en.wikipedia.org/wiki/
Category:Wikipedia_sockpuppets
[6]
https://meta.wikimedia.org/wiki/Research:Understanding_hoax_articles_on_Eng…
--
*Dario Taraborelli *Director, Head of Research, Wikimedia Foundation
wikimediafoundation.org • nitens.org • @readermeter
<http://twitter.com/readermeter>
My recent mail to this list resulted in the following response (quoted
as plain text; the original included images and links)
Please can someone arrange to block maniac92630(a)gmail.com from this
list until this automated tool is deactivated?
On 2 October 2017 at 12:04, <maniac92630(a)gmail.com> wrote:
>
> Hello!
>
> I use a new email filtering service called BitBounce to better filter my spam. To deliver your email to my inbox, please click the link below and pay the small Bitcoin fee. Thank you!
>
> You can sign up or get more info about BitBounce by clicking here.
>
> To deliver your email:
>
> We’ve never met. I’ll pay your fee.
>
> I know you. Add me to your whitelist.
>
> Email that pays.
>
> BitBounce is a product by:
>
> Turing Cloud
>
> BitBounce.com
>
> Redwood City, CA
>
> BitBounce integrates with:
>
> CoinBase
>
> CoinBase.com
>
> San Francisco, CA
--
Andy Mabbett
@pigsonthewing
http://pigsonthewing.org.uk