Az Annif gépi tárgyszavazó rendszer magyarországi adaptációjának feltételei és lehetőségei

Abstract

A recurring annual international event for library information science professionals, the Semantic Web in Libraries (SWIB) conference is a showcase for the latest innovations in linked data and semantic web applications. A regular feature of this conference is the tutorial on the Annif automated subject indexing tool, to which the National Széchényi Library’s Office for GLAM Standards has had the opportunity to delegate a member. Developed by the National Library of Finland, the open source Annif has quickly become one of the flagship for artificial intelligence solutions in libraries. It is already used on routine basis in Finland, more and more libraries are adapting the system, primary in Europe. The secret of its success is both the open source nature of the Annif and the flexibility to integrate with existing components, such as various vocabularies and natural language processing (NLP) algorithms and other AI developments. The Annif therefore is a very useful tool that can significantly facilitate and even revolutionize content discovery, but the key to successful adaptation is the conscious preparation for this task by the library system. It is not only to provide the technical requirements and standardized indexing and cataloging, but also a paradigm shift is needed to enable the current Hungarian library system to adapt such an innovative solution, which brings a new approach to librarians everyday professional life in Hungary. In my paper, summarizing the experiences of the SWIB workshop, I explore these issues and summarize the basic features of the Annif.