TY  - CHAP
U1  - Konferenzveröffentlichung
A1  - Brunner, Annelen
A1  - Tu, Ngoc Duyen Tanja
A1  - Weimer, Lukas
A1  - Jannidis, Fotis
ED  - Ebling, Sarah
ED  - Tuggener, Don
ED  - Hürlimann, Manuela
ED  - Cieliebak, Mark
ED  - Volk, Martin
T1  - To BERT or not to BERT – Comparing contextual embeddings in a deep learning architecture for the automatic recognition of four types of speech, thought and writing representation
T2  - Proceedings of the 5th Swiss Text Analytics Conference (SwissText) & 16th Conference on Natural Language Processing (KONVENS)
N2  - We present recognizers for four very different types of speech, thought and writing representation (STWR) for German texts. The implementation is based on deep learning with two different customized contextual embeddings, namely FLAIR embeddings and BERT embeddings. This paper gives an evaluation of our recognizers with a particular focus on the differences in performance we observed between those two embeddings. FLAIR performed best for direct STWR (F1=0.85), BERT for indirect (F1=0.76) and free indirect (F1=0.59) STWR. For reported STWR, the comparison was inconclusive, but BERT gave the best average results and best individual model (F1=0.60). Our best recognizers, our customized language embeddings and most of our test and training data are freely available and can be found via www.redewiedergabe.de or at github.com/redewiedergabe.
KW  - Einbettung <Linguistik>
KW  - Deutsch
KW  - Testdaten
KW  - Textanalyse
Y1  - 2020
U6  - https://nbn-resolving.org/urn:nbn:de:bsz:mh39-115617
UN  - https://nbn-resolving.org/urn:nbn:de:bsz:mh39-115617
UR  - https://ceur-ws.org/Vol-2624/paper5.pdf
SN  - 1613-0073
SS  - 1613-0073
SP  - 11
S1  - 11
PB  - CEUR-WS
CY  - Aachen
ER  -