We're designing and testing recommender systems to accelerate the rate of content creation and help our contributors and content translators fill the gap between content supply and demand in Wikipedia across languages. Watch a presentation of our results or read more about this project.
We're building machine learning classifiers to make automated quality assessment available as a service to content curators and tool developers. We designed APIs to compute article and edit quality scores and an extensible framework for rolling out new models trained with user input. Learn more about this project and the code behind it. (press)