Project description

The use of innovative methods provided by AI (Artificial Intelligence) for processing and analysing texts and metadata is set to further improve the quality of machine-based subject cataloguing. Promising AI developments suitable for the cataloguing of text-based media works with a highly varied vocabulary are to be analysed, selected, combined and adapted with this aim in mind.

The project explores which AI methods can be used for the machine-based processing and analysis of texts in natural language, in order to receive the most complete and accurate cataloguing data for describing the content. The aim is to provide quality-assured semantic links between the media works and subject headings from the Integrated Authority File (GND).

The new methods are to be provided as flexible, reusable tools (Open Source Tools) for the subject cataloguing of publications in libraries and other institutions with comparable tasks.

Background

Literature researches in the catalogue of the German National Library should lead to complete and reliable results. The subject cataloguing of media works using the authority data of the GND plays an important role in this. The link with GND entities that comprehensively describe the content facilitates definitive search entries and the semantic networking with additional information systems.

The collection of online publications in particular creates a vast quantity of media works requiring cataloguing. The German National Library can only perform the intellectual subject cataloguing for a proportion of the collected media works. This is why it has also been using machine-based cataloguing processes for some years now. The project seeks to find a fundamentally new approach to some of the as-yet-unresolved challenges of machine-based cataloguing.

Network for machine-based cataloguing processes

The project serves the transfer of suitable technologies from the sphere of research and development into practice within the library’s routine activities. The requirements are based around the tasks of the German National Library. A key part of the project is an intensive exchange of information and experience with institutions working in research, development and applications. The Network of machine-based cataloguing processes forms the basis of this exchange.

Project framework

Funding

With the AI strategy, the federal government is supporting the researching, development and application of innovative technologies. The German National Library’s project is being funded by the Minister of State for Culture and the Media.

Duration

26 April 2021 to 31 March 2025

Contact

ki-ee@dnb.de