Historical Document

Research on historical documents focuses on developing computational methods to improve access to and analysis of these valuable resources. Current efforts leverage deep learning models, particularly transformer architectures and GANs, to address challenges such as optical character recognition (OCR) error correction, lacunae restoration, and entity recognition in diverse document types (e.g., newspapers, manuscripts). These advancements facilitate large-scale analysis of historical data, enabling new insights in fields ranging from digital humanities and social sciences to forensic document analysis, ultimately enhancing historical research and knowledge discovery.

Papers