Quantifying Culture: AI and Cultural Heritage Collections

Making cultural heritage more accessible and understandable for both current and future researchers and visitors presents a significant challenge. This project aims to develop a digital model that leverages artificial intelligence (AI) and machine learning (ML) to scan, interpret, and automatically classify cultural objects. In collaboration with the Swedish National Heritage Board, the National Museums of World Culture, the National Archives of Sweden, and Uppsala University Museum (Gustavianum), we are exploring innovative solutions to enhance the accessibility of cultural collections.

The research project “Quantifying Culture: AI and Cultural Heritage Collections” focuses on examining the current applications of AI in cultural heritage and exploring its future potential. The researchers aim to demonstrate how AI can be critically applied to label and interpret archival material within Swedish cultural heritage collections. This technology, similar to facial recognition or automatic translation systems, offers exciting possibilities for making cultural heritage more inclusive and representative of diverse perspectives.

Using 19th and 20th-century archives and collections from our collaborating partners, we are investigating how AI-generated descriptions of visual and textual materials can be tailored to avoid reinforcing standardized narratives from a single viewpoint. For example, when scanning an image of a child in a skirt on a park bench, AI might identify “park bench,” “child,” and “skirt,” and conclude that the child is a girl based on cultural norms. However, in a different cultural context, such as Scotland, a skirt may be traditional boys’ attire. Our goal is to develop more sophisticated mathematical models and algorithms that enable AI to recognize and respect cultural nuances across various communities and regions.

This project brings together innovative methods from the humanities, social sciences, and AI research to explore how AI can address the qualitative aspects of cultural heritage materials. By integrating critical and ethical theories with algorithms and mathematics, we aim to create a more advanced and culturally sensitive AI system that enhances the accessibility and understanding of our shared cultural heritage.