Generating Derivational Morphology with BERT

Valentin Hofmann, Janet B. Pierrehumbert, Hinrich Schütze

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.

Knowledge Graph

arrow_drop_up

Comments

Sign up or login to leave a comment