Charles Explorer logo
🇬🇧

Diacritics Restoration using BERT with Analysis on Czech language

Publication at Faculty of Mathematics and Physics |
2021

Abstract

We propose a new architecture for diacritics restoration based on contextualized embeddings, namely BERT, and we evaluate it on 12 languages with diacritics. Furthermore, we conduct a detailed error analysis on Czech, a morphologically rich language with a high level of diacritization.

Notably, we manually annotate all mispredictions, showing that roughly 44% of them are actually not errors, but either plausible variants (19%), or the system corrections of erroneous data (25%). Finally, we categorize the real errors in detail.

We release the code at https://github.com/ufal/bert-diacritics-restoration.