The Feminist Data Set
On the intersections between art and machine learning, Caroline Sinders is working on an ongoing project to create a feminist data set. Rooted in a collaborative feminist ethics, the project collects “art works, essays, interviews, and books that are from, about or explore feminism and a feminist perspective” to build a larger feminist data set that can then be used to train AI “to locate feminist and other intersectional ways of thinking across digital media distributed online”. Check out the full project description and more images of the project on her website.
The project has been present at various tech/art conferences with participatory workshops or lectures. See for example the workshop description for transmediale (Feb 2019), a festival for arts and digital culture in Berlin, or watch this closely related talk by Caroline Sinders at re:publica 2017 titled ‘Emotional Trauma, Machine Learning and the Internet‘. The 2019 edition of re:publica will see her return with a keynote on the feminist data set and a workshop, I’ll link the video of the talk when it’s published.
While the project has been on my radar for a good while, I’ve not yet had the pleasure of attending one of the workshops in person. You know how it goes, either it’s happening nowhere in reach, or I only find out after it’s already fully booked. I’ll definitely keep keeping an eye out for upcoming instances of the workshop at conferences or events I’m attending!