The National Archives hold a vast collection of historical materials, including, inter alia, all known Swedish medieval letters, administrative national/royal administrative documents from the 16th century and onwards, court records from the 17th century onwards, diplomatic correspondence from the 17th century onwards, and a large collection of historical maps, both general and detailed. These archival records are available online via an archival search service, via APIs, and with select datasets of extracted information from archives available for download.
A recently founded AI-lab at the National Archives has developed and made available models and training data for transcription of Swedish historical handwriting and other machine learning and language models. These resources are valuable for researchers in AI and machine learning focusing on training and evaluating models, as well as for researchers in humanities and social sciences employing digital methods in their work.
To foster collaboration and exploration of these resources, the National Archives, WARA M&L, and WASP-HS the Wallenberg Autonomous System Programme in Humanity and Society, are organizing a series of joint workshops. The first of these will take place on Friday, February 28, from 1:00 to 3:00 PM, and is primarily directed toward researchers in the humanities and social sciences.
The objectives of the workshop are twofold: to provide participants with an understanding of the existing archives and resources and to explore how these assets can be leveraged to support current and future research initiatives.
We are very much looking forward to exploring AI and archives with you.
Please sign up here: https://eu.invajo.com/l/LRhzc56CWD