Wikimedia Scoring Platform team

Wikimedia Foundation project Scoring Platform We build scoring services that enable advanced wiki tools Group: Technology Team members: Staff Chris Albon Andy Craze Kevin Bazira Volunteers Sumit Nate TeBlunthuis Management: Chris Albon

Wikimedia Scoring Platform team logo

Welcome to the home of the Wikimedia Scoring Platform team. For our past work as an ad-hoc, volunteer project, see m:Research:Revision_scoring_as_a_service/Archived. As of July, 2017, we're an officially funded team operating within the Technology Department at the Wikimedia Foundation.

Team

• 

  Chris Albon Director of Machine Learning

• 

  Andy Craze
  Senior Software Engineer

• 

  Kevin Bazira
  Associate Software Engineer

Former

• とある白い猫 (IEG)
• Arthur Tilley (IEG)
• He7d3r (IEG/Volunteer)
• YuviPanda (Volunteer)
• Sumit (Research Intern)
• Ewitch (Research Intern)
• Marius Hoch (Software Engineer)
• Max Klein (Software Engineer)
• Adam Wight (Software Engineer)
• Natalia Timakova (Research Intern)
• Amir Sarabadani (Software Engineer)
• Nate TeBlunthuis (Research Intern)
• James Hare (Associate Product Manager)
• Aaron Halfaker (Principal Research Scientist & Team Lead)
• Habeeb Shopeju (Software Engineering Intern)

Mission

There are two divergent conversations about artificial intelligences—in one, robots will save us from ourselves, and in the other, they destroy us. AI has great potential to help our projects scale by reducing the work that our editors need to do and enhancing the value of our content to readers, but AIs also have the potential to perpetuate biases and silence voices in novel and insidious ways. Imagine a world where AIs are powerful, open, accessible, audit-able tools that Wikimedians use to make their work easier. We develop and maintain AI services (like ORES ) and the related technologies as a means to unlocking that future. We're an experimental, research focused, community supported, AI-as-a-service team. Our work focuses on balancing efficiency and accuracy with transparency, ethics, and fairness.

1. We deliver realtime advanced machine prediction technologies in an easy-to-use format via a web service
2. We focus on supporting all wiki communities who will collaborate with us
3. We develop strategies and technological support for identifying and mitigating hidden biases
4. We publish, communicate, and promote what we've learned so that others can benefit

Projects

• ORES -- Machine learning prediction as a web service (see the list of tools that use ORES)
• m:Wiki labels -- Training interface where Wikipedians teach machines how to perform important tasks
• revscoring -- A machine prediction "scoring" framework for building prediction models used by ORES
• JADE -- Robust false-positive and feedback gathering system, to allow human refutation and review of ORES scoring.

ORES, machine learning prediction as a web service platform

Wiki labels, a training interface where Wikipedians teach machines how to perform important tasks

revscoring, a machine prediction "scoring" framework for building prediction models used by ORES

Research collaborations

We are working in collaboration with the following research projects studying ORES and the use of AIs in Wikipedia.

• Research:Applying Value-Sensitive Algorithm Design to ORES
• Research:Exploring systematic bias in ORES

Contact us

We're an open project team. There are many ways you can get in contact with us.

• Email: calbon@wikimedia.org
• Mailing list: ai@lists.wikimedia.org
• IRC: #wikimedia-ai ^connect
• Phabricator: #Scoring-platform-team
• Flow board: Talk:Wikimedia Scoring Platform team