DagoBERT: Generating Derivational Morphology with a Pretrained Language Model

Valentin Hofmann, Janet Pierrehumbert, Hinrich Schütze


Abstract
Can pretrained language models (PLMs) generate derivationally complex words? We present the first study investigating this question, taking BERT as the example PLM. We examine BERT’s derivational capabilities in different settings, ranging from using the unmodified pretrained model to full finetuning. Our best model, DagoBERT (Derivationally and generatively optimized BERT), clearly outperforms the previous state of the art in derivation generation (DG). Furthermore, our experiments show that the input segmentation crucially impacts BERT’s derivational knowledge, suggesting that the performance of PLMs could be further improved if a morphologically informed vocabulary of units were used.
Anthology ID:
2020.emnlp-main.316
Volume:
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Month:
November
Year:
2020
Address:
Online
Editors:
Bonnie Webber, Trevor Cohn, Yulan He, Yang Liu
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3848–3861
Language:
URL:
https://aclanthology.org/2020.emnlp-main.316
DOI:
10.18653/v1/2020.emnlp-main.316
Bibkey:
Cite (ACL):
Valentin Hofmann, Janet Pierrehumbert, and Hinrich Schütze. 2020. DagoBERT: Generating Derivational Morphology with a Pretrained Language Model. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 3848–3861, Online. Association for Computational Linguistics.
Cite (Informal):
DagoBERT: Generating Derivational Morphology with a Pretrained Language Model (Hofmann et al., EMNLP 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.emnlp-main.316.pdf
Video:
 https://slideslive.com/38939116
Code
 valentinhofmann/dagobert