Adapting Coreference Algorithms to German Fairy Tales
David Schmidt, Markus K. Krug, Frank Puppe · Zenodo (CERN European Organization for Nuclear Research) · 2022
Coreference Resolution is an important task in natural language processing that enables the combination of locally extracted information on a document level context or even across different documents, e.g. to display the relations between characters of a novel in a character network. In this work, we adapt a rule-based and an end-to-end deep learning algorithm which have previously been used on German novel fragments to the domain of German fairy tales and examine their performance. We find that the adaptation can improve their performance considerably, especially that of the deep learning algorithm. Ein Betrag zur 8. Tagung des Verbands "Digital Humanities im deutschsprachigen Raum" - DHd 2022 Kulturen des digitalen Gedächtnisses.