Algorithmic Equity: A Data-Driven Method for Diversifying Museum Collections
Brigitta Hong
Saint Paul Preparatory Seoul, Seoul, South Korea
Publication date: November 20, 2025
Saint Paul Preparatory Seoul, Seoul, South Korea
Publication date: November 20, 2025
DOI: http://doi.org/10.34614/JIYRC2025II45
ABSTRACT
This research explores how algorithmic tools can promote curatorial equity by identifying underrepresented women artists for inclusion in major museum collections. Focusing on the Art Institute of Chicago (AIC), the study developed a soft-matching algorithm that compares artist traits—medium, gender, demographic background, and classification—with metadata sampled from 1,000 artworks accessed through the AIC’s public API. A curated list of 22 women artists from historically marginalized regions was scored based on their alignment with the museum’s holdings. Results revealed both strong matches and areas of significant underrepresentation, particularly in media like performance and video, and in regions such as Sub-Saharan Africa and South Asia. Conducted using Python in Google Colab, this research demonstrates how computational thinking can support more inclusive curatorial practices. The methodology is scalable to other institutions and offers a transparent framework for using data science to inform diversity-driven acquisition strategies.
This research explores how algorithmic tools can promote curatorial equity by identifying underrepresented women artists for inclusion in major museum collections. Focusing on the Art Institute of Chicago (AIC), the study developed a soft-matching algorithm that compares artist traits—medium, gender, demographic background, and classification—with metadata sampled from 1,000 artworks accessed through the AIC’s public API. A curated list of 22 women artists from historically marginalized regions was scored based on their alignment with the museum’s holdings. Results revealed both strong matches and areas of significant underrepresentation, particularly in media like performance and video, and in regions such as Sub-Saharan Africa and South Asia. Conducted using Python in Google Colab, this research demonstrates how computational thinking can support more inclusive curatorial practices. The methodology is scalable to other institutions and offers a transparent framework for using data science to inform diversity-driven acquisition strategies.
