Shorthand, long game

The University of Pittsburgh holds numerous manuscripts written in Stolze-Schrey shorthand, presented in our Digital Collections system, and stored as images in our Digital Preservation system. While this writing system was once common, current researchers are often at a loss when encountering these primary sources, and this disconnect will likely increase over time. The University Library System is coordinating resources across information technology, archives, language studies, and shorthand practitioners to train a machine learning model to recognize the handwriting system. We'll use this model to perform handwriting text recognition to capture this content to our Digital Collections and Digital Preservation systems. For this presentation, I will describe the motivations, methodology, and preservation output of this work-in-progress to ensure that this shorthand is accessible for the long term.