This month, our research showcase
<https://www.mediawiki.org/wiki/Wikimedia_Research/Showcase#March_2016> hosts
Andrei Rizoiu (Australian National University) to talk about his work
<http://cm.cecs.anu.edu.au/post/wikiprivacy/> on *how private traits of
Wikipedia editors can be exposed from public data* (such as edit histories)
using off-the-shelf machine learning techniques. (abstract below)
If you're interested in learning what the combination of machine learning
and public data mean for privacy and surveillance, come and join us
this *Wednesday
March 16* at *1pm Pacific Time*.
The event will be recorded and publicly streamed
<https://www.youtube.com/watch?v=Xle0oOFCNnk>. As usual, we will be hosting
the conversation with the speaker and Q&A on the #wikimedia-research
channel on IRC.
Looking forward to seeing you there,
Dario
Evolution of Privacy Loss in WikipediaThe cumulative effect of collective
online participation has an important and adverse impact on individual
privacy. As an online system evolves over time, new digital traces of
individual behavior may uncover previously hidden statistical links between
an individual’s past actions and her private traits. To quantify this
effect, we analyze the evolution of individual privacy loss by studying the
edit history of Wikipedia over 13 years, including more than 117,523
different users performing 188,805,088 edits. We trace each Wikipedia’s
contributor using apparently harmless features, such as the number of edits
performed on predefined broad categories in a given time period (e.g.
Mathematics, Culture or Nature). We show that even at this unspecific level
of behavior description, it is possible to use off-the-shelf machine
learning algorithms to uncover usually undisclosed personal traits, such as
gender, religion or education. We provide empirical evidence that the
prediction accuracy for almost all private traits consistently improves
over time. Surprisingly, the prediction performance for users who stopped
editing after a given time still improves. The activities performed by new
users seem to have contributed more to this effect than additional
activities from existing (but still active) users. Insights from this work
should help users, system designers, and policy makers understand and make
long-term design choices in online content creation systems.
*Dario Taraborelli *Head of Research, Wikimedia Foundation
wikimediafoundation.org •
nitens.org • @readermeter
<http://twitter.com/readermeter>