Revisiting Neural Language Modelling with Syllables
This work addresses the problem of efficient language modeling for multilingual applications, though it is incremental by revisiting an underused unit.
The paper tackled language modeling by using syllables as units, showing that they outperform characters, annotated morphemes, and unsupervised subwords with comparable perplexity in an open-vocabulary generation task across 20 languages.
Language modelling is regularly analysed at word, subword or character units, but syllables are seldom used. Syllables provide shorter sequences than characters, they can be extracted with rules, and their segmentation typically requires less specialised effort than identifying morphemes. We reconsider syllables for an open-vocabulary generation task in 20 languages. We use rule-based syllabification methods for five languages and address the rest with a hyphenation tool, which behaviour as syllable proxy is validated. With a comparable perplexity, we show that syllables outperform characters, annotated morphemes and unsupervised subwords. Finally, we also study the overlapping of syllables concerning other subword pieces and discuss some limitations and opportunities.