Submissions:2019/Discussing AI and Machine Learning in the Wikimedia movement

From WikiConference North America
Revision as of 04:12, 23 September 2019 by Fuzheado (talk | contribs) (Created page with "{{WCNA 2019 Session Submission |theme=Reliability of Information<br />+ Tech & Tools<br /> |type=Round Table |abstract=Since Wikipedia's founding in 2001, the state of the art...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

This submission has been noted and is pending review for WikiConference North America 2019.


Discussing AI and Machine Learning in the Wikimedia movement


Reliability of Information
+ Tech & Tools

Type of session:

Round Table


Since Wikipedia's founding in 2001, the state of the art in artificial intelligence and machine learning has advanced dramatically. Today, there are numerous providers of AI as a service, or high quality machine learning platforms that anyone can utilize to perform high quality work at scale. Wikimedia's highest profile engagement with machine learning can be seen in the ORES (Objective Revision Evaluation Service) project, as a way to automatically generate article quality ratings and to detect different types of user behavior such as vandalism and poor edits, greatly aiding in the ability to reject bad behavior. It is also the basis of the Wiki Education Foundation dashboard and the Programs & Events Dashboard to evaluate article progress and editor impact.

AI is used not only in the lexical space to parse text or generate spoken content on voice assistants like Alexa, Siri, and Google Home, but also in the visual domain for remarkable applications related to computer vision. High quality image recognition and classification has become so common that the Wikimedia Foundation has already put in plans to provide machine-generated tags for Wikimedia Commons on uploads so that users can help confirm this auto-created metadata. In the area of generated content, the creation by some parties of uncannily good AI-generated faces (GANs) under public domain licenses pose tough questions about their appropriateness, scope and use in Wikimedia projects.

There are enough concerns about the future of AI and machine learning within the Wikimedia ecosystem that this would be a great discussion topic as we imagine how we might think of innovative uses, and where we might tread more carefully.

Ideally we would have the participation of Wikimedia Foundation staff involved with AI efforts in addition to third party and community members currently doing work with these technologies.

Academic Peer Review option:


Author name:

Andrew Lih

E-mail address:

Wikimedia username:


Affiliated organization(s):

Wikimedia DC

Estimated time:

45 minutes

Preferred room size:


Special requests:

Have you presented on this topic previously? If yes, where/when?:


If your submission is not accepted, would you be open to presenting your topic in another part of the program? (e.g. lightning talk or unconference session)

Lightning or unconference