Large Collection

Research on large collections focuses on efficiently processing and analyzing massive datasets across diverse domains, from book catalogs and census records to human motion and time series data. Current efforts leverage techniques like many-to-many matching algorithms, temporal matrix factorization, and deep learning models (e.g., BERT, CLIP) to extract meaningful information and build predictive models. These advancements enable improved data management, facilitate large-scale analyses across various fields, and offer significant potential for enhancing applications ranging from cultural heritage preservation to economic forecasting.