DE-BIAS

For decades, European archives, museums, and other cultural heritage institutions have been cataloging and describing their collections of historical objects. However, existing catalog data, often created from the perspective of the majority society, is rarely updated and does not reflect evolving language and societal changes. As a result, these records may contain outdated views and terms that can be perceived as offensive or hurtful. Yet the vast size of many collections and the complexity involved make updating this catalog data challenging.
To support a more inclusive, contemporary, and feasible approach to cataloging cultural collections, the DE-BIAS project is developing an AI-powered tool to automatically detect discriminatory and outdated terms, providing suggestions for respectful, current alternatives. This tool is based on a vocabulary being developed around key themes: 1. migration and colonial history, 2. gender and sexual identity, and 3. ethnicity and ethno-religious identity. In this effort, the project actively engages minority communities in the vocabulary’s creation to ensure their perspectives are included and respected. Through this collaboration, DE-BIAS aims to enhance representation and participation from these communities.
The tool will be integrated into Europeana, the European search portal for archival collections from over 5,000 archives, museums, and related institutions, and made available as a standalone version for individual institutions. This will allow them to analyze their datasets and automatically identify problematic expressions. Lastly, the project will provide training resources and handouts to ensure the long-term adoption and impact of these advancements within cultural institutions.