Fully Character-Level Neural Machine Translation without Explicit Segmentation

TL;DR

This paper presents a character-level neural machine translation model that eliminates the need for explicit segmentation, achieving competitive performance and enabling effective multilingual translation sharing.

Contribution

It introduces a character-level convolutional encoder that reduces sequence length and enables training at speeds comparable to subword models, outperforming them in multilingual settings.

Findings

Outperforms subword models on WMT'15 DE-EN and CS-EN

Achieves comparable results on FI-EN and RU-EN

Multilingual character-level model surpasses language-specific models in BLEU and human judgment

Abstract

Most existing machine translation systems operate at the level of words, relying on explicit segmentation to extract tokens. We introduce a neural machine translation (NMT) model that maps a source character sequence to a target character sequence without any segmentation. We employ a character-level convolutional network with max-pooling at the encoder to reduce the length of source representation, allowing the model to be trained at a speed comparable to subword-level models while capturing local regularities. Our character-to-character model outperforms a recently proposed baseline with a subword-level encoder on WMT'15 DE-EN and CS-EN, and gives comparable performance on FI-EN and RU-EN. We then demonstrate that it is possible to share a single character-level encoder across multiple languages by training a model on a many-to-one translation task. In this multilingual setting, the character-level encoder significantly outperforms the subword-level encoder on all the language pairs. We observe that on CS-EN, FI-EN and RU-EN, the quality of the multilingual character-level translation even surpasses the models specifically trained on that language pair alone, both in terms of BLEU score and human judgment.