Revolucionar el acceso al patrimonio librario: los sistemas de HTR entre Humanidades Digitales y ciencia de la información

The present work aims to offer a state of the art on recent developments in the field of automatic transcription of historical printed documents and manuscripts with HTR (Handwritten Text Recognition) systems, focusing primarily on the recent creation of HTR general models. In this regard, the main characteristics of the most widespread tools and the workflow for generating text recognition models are explained. Secondly, a significant sample of the models currently available is provided, insisting on the production process, the criteria adopted and the evaluation of the results, in relation to the experience matured by the Progetto Mambrino research group of the University of Verona. Finally, some future research directions are provided for the creation and dissemination of these resources, emphasizing the need to seek greater synergy between the academic context, computer experts and memory institutions.