Editors’ Summary: This piece provides a detailed report on using AI to transcribe manuscripts and the development of standards for transcription of medieval manuscripts. The research team began by creating a uniform database of over 300 medieval manuscripts across many languages. The standardization of this database allowed the corpus to be used to train an AI model. This model was then used to transcribe thousands of medieval manuscripts. This piece is a useful roadmap on best practices for using AI models for transcription.
Source: CoMMA: thousands of medieval manuscripts finally transcribed