Thanks to UWB researchers' technology, people can search online in cross-border information sources

Cooperation Press Release Science

The Czech-Bavarian Porta Fontium website, which since 2011 has been digitizing archives documenting the coexistence of Czechs and Germans in the cross-border region, has expanded the database of available archival information sources.

Thanks to the cooperation between researchers from the Faculty of Applied Sciences (FAV) of the University of West Bohemia, archivists from the State Regional Archive in Pilsen, and the Bavarian State Archives and historians, both the scientific community as well as the general public can access online not only registers of births and deaths, chronicles, official documents, official records and periodicals, but also 7,800 historical maps and plans dating from before 1918. The portal's new feature is dual-language search, including location details.

The Modern Access to Historical Sources 2018-2021 project, supported from the Czech Republic – the Free State of Bavaria cross-border cooperation program, is based on one-of-a-kind collaboration between experts from Czechia and Bavaria in the fields of history, historical building surveys, archival science and information technology. This cooperation made possible the expansion of the Porta Fontium portal with new types of archival documents, such as maps and plans, as well as the possibility to effectively acquire information from these and other graphic documents. Thanks to the technology of automated processing of printed and hand-written documents developed by FAV experts, it is possible to search in all types of information sources, including maps, scanned hand-written texts and photographs.

"Earlier searches through the portal relied on manual sifting and reading through extensive hand-written records. Nowadays, users can take advantage of the dual-language full-text search," explains Pavel Král from the Natural Language Processing Group in the Department of Computer Science and Engineering at the New Technologies for the Information Society Research Center of the Faculty of Applied Sciences, whose team of five researchers, made up of Ladislav Lenc, Jiří Martínek, Martin Prantl and Josef Baloun from the Faculty of Applied Sciences, developed and integrated into the portal new tools for searching and viewing texts and graphic documents.

"The portal is now also connected with map data, so it is possible not only to search through maps themselves, but also various types of documents based on localization data, i.e., the name of the town or place which they relate to. The selected documents can be viewed directly on the map," Pavel Král points out the new functionality. To make historical documents accessible, the latest knowledge in the area of natural language processing and computer vision was used, in combination with new methods of deep learning.

Using artificial intelligence, IT experts made orientation within all types of information sources easier for historians, regional researchers, genealogists as well as the general public, giving them the possibility to carry out extensive research.

"Selected archival documents from both sides of the border which as a result of historical events were forcibly divided, including historical maps and plans of the cross-border region from before 1918, are now available online in one place. Thanks to this online resource, their originals are protected," says Karel Halla from the State Regional Archive in Pilsen.

Apart from IT experts from the Faculty of Applied Sciences, the project involved the participation of archivists from the State Regional Archive in Pilsen and historians from the West Bohemia Institute for the Protection and Documentation of Heritage Sites. On the Bavarian side, the parties involved were archivists from the Directorate General of the Bavarian State Archives (Generaldirektion der Staatlichen Archive Bayerns) and researchers from the University of Erlangen-Nuremberg (Friedrich-Alexander-Universität Erlangen-Nürnberg).



Faculty of Applied Sciences

Šárka Stará

11. 05. 2022