Nebraska Forum on Digital Humanities

From Exclusion to Inclusion: Creating Responsible and Equitable Information and AI Systems for Digital Cultural Heritage

Sponsored by the Angle Chair in the Humanities, the School of Computing, the Department of Mathematics, the School of Global Integrative Studies, the Office of Research and Economic Development Grand Challenges Seed Grant, and the Center for Digital Research in the Humanities.

Forum co-organizers: Scout Calvert (University Libraries), Carrie Heitman (School of Global Integrative Studies), Yi Liu (Postdoctoral Researcher), and Leen-Kiat Soh (School of Computing)

Legacy systems for organizing knowledge and cultural heritage materials embed biases that distort the historical and scientific record, in turn biasing knowledge produced from these sources. The need for new technologies to efficiently describe and retrieve these materials is tied to the need to understand how distortions get built into knowledge systems over time and space. Machine learning techniques promise the potential to automate aspects of description and retrieval, but also amplify bias and distortion in many applications, replicating and turbocharging the worldviews of their creators and concretizing selected epistemologies that reinforce the status quo.

What would it take to develop machine learning for the tasks of managing cultural heritage materials—ranging from texts to images to videos to data—while generating radically new methods for re-comprehending research materials, allowing radically novel questions to be asked, embracing and foregrounding multiplicity of cultural and knowledge systems, addressing racism and bias in the collection, description, conceptualization and use of these materials, and building capacity for critical understanding of these technological and social systems at all levels?

This final question is at the heart of this year’s forum.