Context Models for OOV Word Translation in Low-Resource Languages
This addresses translation issues for low-resource languages, but it is incremental as it builds on existing methods like neural MT with context models.
The paper tackled the problem of out-of-vocabulary word translation in low-resource languages by evaluating target-language context models, showing that neural language models with additional context beyond the current sentence are most effective for disambiguation, and demonstrated performance improvements over state-of-the-art neural MT systems in five out of six low-resource language pairs.
Out-of-vocabulary word translation is a major problem for the translation of low-resource languages that suffer from a lack of parallel training data. This paper evaluates the contributions of target-language context models towards the translation of OOV words, specifically in those cases where OOV translations are derived from external knowledge sources, such as dictionaries. We develop both neural and non-neural context models and evaluate them within both phrase-based and self-attention based neural machine translation systems. Our results show that neural language models that integrate additional context beyond the current sentence are the most effective in disambiguating possible OOV word translations. We present an efficient second-pass lattice-rescoring method for wide-context neural language models and demonstrate performance improvements over state-of-the-art self-attention based neural MT systems in five out of six low-resource language pairs.