1The author of this paper.
2See https://github.com/NatLibFi/Annif/wiki/Backend%3A-Omikuji for its integration into Annif.
3See https://github.com/NatLibFi/Annif/wiki/Backend%3A-fastText for its integration into Annif.
4See https://github.com/NatLibFi/Annif/wiki/Backend%3A-STWFSA for its integration into Annif.
5Experiments in the AutoSE context have shown that author keywords improve the F1 score (
6For example, there had been a regular exchange of ideas prior to the German National Library launching their own Annif-based “cataloguing machine” and associated AI project, see
7The number for intellectual subject indexing at ZBW is around 30,000 records per year.
8∼147,000 in 2021; ∼500,000 in 2020.
9For a short description see
10(January 2023) A prototypical version exists – the next step is to launch the UI internally.
11In
12The goal is to increase the so-called
13“Technology transfer is a bilateral process between sender and receiver: there is a ‚process of feed-back‘ from the sender to the receiver, which allows interested parties to obtain more information (knowledge) on the use of technology transferred” (
14In terms of the challenges compiled by