At the beginning of 2022, the implementation of the ASARP project – Automated system of analytical scheduling of periodicals focused on the development of a system for automation and cataloging of digitized periodicals at the level of articles using machine learning.
In cooperation with the members of the consortium, a software solution will be created, which will enable the enrichment of the digital library with another function in the form of created records of the analytical schedule. Memory institutions (and especially libraries) have a rich source of digitized periodicals, which is not used. The project partner – the Moravian Regional Library – has a collection of several hundred thousand articles, which are converted into digital form, but are not fully cataloged, therefore, it is difficult for users to effectively find relevant results. There is an accessibility issue with digitized library content. Although there is a lot of digitized material, it is very difficult to access for researchers. The content and annotations are basically hidden in the text, it is difficult to find in it. In addition, individual years are often bundled into one pdf file. Although this pdf file can be searched in full text, due to the number of articles and texts it is inefficient. The only current tool for indexing articles at the analytical breakdown level is manual processing, which is hugely time consuming (and therefore costly). For this reason, this step in digitizing periodicals is neglected and hardly done. Due to their properties, output quality and adaptability, current tools are not suitable for processing large library collections. Therefore, a way is being sought for the permanent automation of the creation of the analytical schedule of periodicals.
The aim of the project is to combine an efficient user interface and the latest machine learning methods and create software usable by memory institutions for processing and categorization of collections of digitized periodicals, which will replace current tools with their features, output quality and adaptability unsuitable for processing large library collections due to the huge need for manual work.
Project implementation period: 1 January 2022 – 31 May 2023.
Total eligible project expenses: CZK 10,960,792.
Project partners: Cosmotron Bohemia, s.r.o., Moravian Regional Library in Brno.
The project is funded by the European Union through the Operational Program Enterprise and Innovation for Competitiveness.
