Kaleidoscopic Indexing

As generative AI transforms access to historical knowledge, existing indexing methods fail to capture the interpretive nuance, emotional depth, and cultural complexity of archival materials. This project develops "kaleidoscopic indexing," a humanistically-informed AI framework for interpretive access to multilingual and multimodal Holocaust archives. By integrating humanistic heuristics—such as cultural and historical context, narrative expressivity, ambiguity, nuance, absence, and perspective—into open-source tools, we prototype a new model for ethically-grounded archival interpretation.