What are LAC’s intentions on Handwritten Text Recognition technology?

“Handwritten Text Recognition (HTR) technology is now a mature machine learning tool, becoming integrated in the digitisation processes of libraries and archives speeding up the transcription of primary sources and facilitating full text searching and analysis of historic texts at scale.”

That’s a quote from the 2022 journal article Understanding the application of handwritten text recognition technology in heritage contexts: a systematic review of Transkribus in published research.

If there was any doubt, the successful partial indexing of the 1931 census of Canada by Ancestry, using Library and Archives Canada images, as well as their earlier indexing of the 1950 US census, confirms the capability.

What are LAC’s plans? I turned to the recently published Vision 2030: Discover. Understand. Connect — A strategic plan to 2030.

“While the collections remain at the centre of LAC’s identity, the ability of Canadians to discover, understand and connect with the collections frames our vision of service. Vision 2030 will aim to provide access to the collections to anyone and from anywhere, as well as a framework to identify and respond quickly to users’ needs.”

What is the reality of “provide access?” Does that mean online? Why “a framework” and not “a system”?

Vision 2023 includes one mention of AI

“LAC will benefit from work being conducted in other domains, such as the digital humanities and ethical artificial intelligence.”

Am I alone in reading this as LAC having no plans to implement the mature machine learning tool of  Handwritten Text Recognition (HTR) technology in its own operations to facilitate easy access to documentary heritage, as with the 1931 census, and why not?



One Reply to “What are LAC’s intentions on Handwritten Text Recognition technology?”

Leave a Reply

Your email address will not be published. Required fields are marked *