Character-based Neural Machine Translation
This work addresses a key bottleneck in neural machine translation for languages with complex morphology, offering an incremental improvement over existing methods.
The paper tackled the challenge of large vocabularies and morphologically rich languages in neural machine translation by proposing a character-based embedding system with convolutional and highway layers, resulting in improvements of up to 3 BLEU points on the German-English WMT task.
Neural Machine Translation (MT) has reached state-of-the-art results. However, one of the main challenges that neural MT still faces is dealing with very large vocabularies and morphologically rich languages. In this paper, we propose a neural MT system using character-based embeddings in combination with convolutional and highway layers to replace the standard lookup-based word representations. The resulting unlimited-vocabulary and affix-aware source word embeddings are tested in a state-of-the-art neural MT based on an attention-based bidirectional recurrent neural network. The proposed MT scheme provides improved results even when the source language is not morphologically rich. Improvements up to 3 BLEU points are obtained in the German-English WMT task.