Computational Analysis of Early Printed Book Descriptions

The British Museum's Catalogue of Books Printed in the XVth Century is a treasure trove of provenance, binding, and dating information — locked inside a complex, variable layout that defeats standard OCR. This PhD project applies Transkribus field models, custom code, and human review to volume XI (English incunabula), extracting structured data that will enrich over 200 entries in the MEI, ISTC, and British Library catalogues. A practical demonstration of how hybrid AI-human workflows can unlock specialist bibliographic resources for the wider research community.