About


In brief

On this platform you can perform full text searches and browse automatically transcribed documents from 19th century Finnish Renovated District Court Records.

Material

At the moment Finnish Renovated District Court Records from the years 1809–1870 are available on the platform. As an exception, the court districts from the parts of Karelia that were ceded to the Soviet Union are not included. These documents are held at the National Archives’ Mikkeli office as a part of Viipuri hovioikeus (Court of Appeal) archives. The excluded districts are Jääski, Ranta/Stranda and Äyräpää districts (1817–) and their predecessor Viipurin kihlakunta/Viborg härad (1812–1816) on the Karelian Isthmus; as well as Salmi and Sortavala (1832–) and Käkisalmi/Kexholm (1863–) districts and their predecessors Käkisalmen pohjoinen (1812–1831), Käkisalmen keskinen (1817–1862) and Käkisalmen eteläinen (1812–1862) in Ladoga Karelia.

The Renovated District Court Records is one of the largest collections in the National Archives of Finland. The collection starts from 1623 when the lower courts were obligated to produce a copy of their court records to the newly founded Turku court of appeals, and continues until the 1970s. The 19th century Renovated Court Records are divided into two series: the main records and the notification records. The material on this platform consists of the notification records, which include registration of deeds, mortgages, traditional life-annuity and guardianship cases. Notification records can for example be used to research family history and track ownership of property.

On the Material page you can find more information about the court records and view maps of the court districts in 19th century Finland.

Technology

The transcriptions on this platform have been generated with so called HTR (Handwritten Text Recognition) technology. HTR uses artificial neural network technology to generate a text recognition model based on human transcribed text. The National Archives has trained a specific model that can read 19th Century court records, and the transcriptions have been generated automatically with this HTR model. Because all the transcriptions are made with a single recognition model, the transcriptions on some pages have more errors than on other pages. Especially the pages from the beginning of the century (before 1830s-1840s) might have more errors in the transcription.

Searching the documents is based on keyword spotting technology (KWS) which searches your query from the confidence matrices that are based on the HTR processed image. In other words, the queried words are not searched simply from the visible text transcription but instead from the information behind the images. The HTR model assigns each letter in the alphabet a value of how certain the model is that that letter corresponds with that part of the image (i.e. a section of a word on a row) when you run the model on a page. When you search for a word, the confidence of your results are based on those confidence values. If you search for single words, you can also adjust the level of confidence. Depending on the confidence threshold of your search, with KWS you might be able to find your query even on pages where the transcription is almost illegible.

Background

This platform was created by the National Archives of Finland within the Horizon 2020 project READ (Recognition and Enrichment of Archival Documents). The project, funded by the European Union, started in January 2016 and ended in June 2019. READ project was an international collaboration between 14 partners drawn from the domains of computer science, archives and humanities research. The aim of the project was to develop handwritten text recognition and keyword spotting technologies.

Now the work continues in the READ-COOP, a European cooperative Society. READ-COOP will serve as the basis for sustaining and further developing platforms, services and tools used in the process of recognition and enrichment of archival documents. At the National Archives work has also continued in the project Making a Modern Archive, in which technologies developed in the READ project have been implemented into the National Archives’ digital infrastructure. The project received funding from the Finnish Ministry of Education and Culture.