Conference dates: November 1st-5th, 2021
As Platinum Sponsor of the 2021 Charleston Conference, with Adam Matthew Digital, we are delighted to confirm our Innovation Lightning Talk session with Quartex partner, Darryl Stuhr, from Baylor University.
Details of date and time to follow in due course.
Harnessing Handwritten Text Recognition technology to improve access and discovery across digital collections
The Armstrong Browning Library on the Baylor University campus is a 19th-century research center dedicated to the study of the lives and works of Victorian poets Robert and Elizabeth Barrett Browning and houses the world's largest collection of Browning material and other fine collections of rare 19th-century books, manuscripts, and works of art.
The correspondence of Elizabeth Barrett and Robert Browning is presented in The Browning Letters collection, which has been the subject of a manual transcription programme running over many years. Researchers can now fully explore this collection and benefit from being able to search and view full text, 100% accurate transcriptions.
In addition to its unparalleled collection of materials relating to the Brownings, the Armstrong Browning Library houses a collection of over 3,000 Victorian letters and manuscripts connected to other prominent and lesser-known British and American figures. The collection includes letters and manuscripts from many notable nineteenth-century authors such as Charles Dickens, William Wordsworth, Samuel Taylor Coleridge, Thomas Carlyle, John Henry Newman, George MacDonald, and John Ruskin, as well as letters and manuscripts from political figures, religious leaders, scientists, artists, art collectors, and explorers.
Transcriptions are not yet available for this collection but the team at Baylor University Libraries is now exploring the use of new technology as part of a project to establish an efficient workflow to deliver accurate transcriptions of these letters. They are running HTR across these manuscripts, without having to retrain the technology to accurately recognise different hands, establishing base accuracy levels across a sample of automated transcriptions, and manually completing transcriptions to achieve 100% accurate outputs.