Generating Derivational Morphology with BERT

05/02/2020

∙

by   Valentin Hofmann, et al.

∙

0

∙

share

Can BERT generate derivationally complex words? We present the first study investigating this question. We find that BERT with a derivational classification layer outperforms an LSTM-based model. Furthermore, our experiments show that the input segmentation crucially impacts BERT's derivational knowledge, both during training and inference.