Handwritten Text Recognition provides a vital case study for the impact of Artificial Intelligence and Machine Learning on libraries. The first part of this online talk will provide a broader context to the discussion by exploring what it means for AI adoption which must be grounded in the values and priorities of the sector. In doing so, it will discuss how the centralisation of technological development has affected libraries use of AI tools, and engage in what it means for libraries to be transparent and ethical actors in the AI space.

The latter part of this talk will use a case study of the National Library of Scotland to provide a practical understanding of how AI-enabled automated transcription can meet the urgent need of providing elemental content through library online services. Hurdles in recognising handwritten and printed materials are highlighted, limiting the accuracy of transcription outputs, with shortcomings discussed and remedies given. This work is shown to be integral to the outreach strategies of the NLS, demonstrating how digital approaches can be at the core of library operations. A picture is provided for how to take up HTR beyond an individual level, to ensure that digital outputs represent a variety of collection users, views and experiences.

Paul Gooding is Professor of Library Studies and Digital Scholarship at the University of Glasgow. His research focuses on understanding the impact of digital library collections on institutions and users, and how library and archival collections can be harnessed for innovative reuse in the Digital Humanities. In recent years his research has explored the use of Artificial Intelligence and Machine Learning in libraries, with a particular interest in how we can make AI tools more transparent and understandable for library users.

Joe Nockels is a PhD researcher at the University of Edinburgh and Glasgow, funded by the Scottish Graduate School for Arts and Humanities (SGSAH) and National Library of Scotland (NLS). He focuses on handwritten text recognition (HTR) from an institutional and user perspective, as well as its implications for the historical method and wider information environment.

Book tickets here