This paper discusses the definition of bias in chcs and other datasets, exploring how it stems from training data and insufficient humanities expertise in generative platforms. As part of the congruence engine project, i am exploring how technology choices affect curatorial as well as research processes and introduce biases in digital cultural heritage and research practice. Library of america hub network to understand the landscape of digital cultural heritage in the us and in connecticut in particular This paper presents observations on digital cultural her Tage in the us and a more specific analysis of the ctda and offers. Table 2 summarizes prominent benchmark datasets employed in cultural bias research, including their composition, cultural representation, and applicability to heritage documentation tasks.
Much of our technology and many or our practices are inherently biased Including search algorithms and metadata systems and tools We must ensure our technology supports multiple cultural viewpoints in intellectual structure, metadata standards and practices, search algorithms and more.
OPEN
