Accessible archives with AI and Citizen science
Karl-Magnus Johansson, The National Archives of Sweden Open Science in the Swedish Context, May 7th 2026
Handwritten Text Recognition (HTR)
Supervised machine learning for automatic transcription of handwritten text from image to text data
Gothenburg Police, Detective Section
Police reports (1868–1902) and telegram messages (1865–1903)
Appr. 25 000 pages
46 bound volumes
A project in collaboration with GPS400: Center for collaborative visual research, University of Gothenburg
Photo: Axel Leonard Svensson, 1901
(Lena Carlssons bildarkiv, The Swedish National Archives, Gothenburg)
Photo: Axel Leonard Svensson, 1900
(Lena Carlssons bildarkiv, The Swedish National Archives, Gothenburg)
Hungerkravallerna
Hungerkravallerna
Studying the epistemic culture that developed in the project
How do volunteers value their participation?
"The public will benefit greatly from this in the future. Those who are not so used to studying old archives will have a much easier time accessing the material.“
”I am learning so much. I had no idea about some local historical details but the project has given me new perspectives on it. When I am out walking with friends I usually tell them about cases that have taken place in different places we visit. I have almost like guided tours.”
“I tell all my friends: 'I'm doing something really fun', and they get so jealous of me, especially during Covid.”
“Projects like this show just how incredibly much there is to discover [in the archives], so many different topics and so on. The project points you in a direction, and you get some free knowledge, and then you contribute, and you learn […] What I like is that – you might already have an interest, and then you get nudged, and suddenly it becomes interesting, or maybe at first you think “oh well,” but the more you learn, the more interesting it gets!
Citizen Science is based on relationships and mutual trust
Glöckner, Murre
Thank you for listening!
Contact: karl-magnus.johansson@riksarkivet.se