# Modeling Target-Side Inflection in Neural Machine Translation

arXiv: https://arxiv.org/abs/1707.06012 · HTML: https://mlsanity.com/paper/1707.06012/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Jul 2017.

## Built on by (most cited first, 9 of 9)

- [Survey of Low-Resource Machine Translation](https://mlsanity.com/paper/2109.00486/) (2021)
- [Isochrony-Aware Neural Machine Translation for Automatic Dubbing](https://mlsanity.com/paper/2112.08548/) (2021)
- [How Suitable Are Subword Segmentation Strategies for Translating Non-Concatenative Morphology?](https://mlsanity.com/paper/2109.01100/) (2021)
- [Morphological Processing of Low-Resource Languages: Where We Are and What's Next](https://mlsanity.com/paper/2203.08909/) (2022)
- [Rule-based Morphological Inflection Improves Neural Terminology Translation](https://mlsanity.com/paper/2109.04620/) (2021)
- [Recent advancements in computational morphology : A comprehensive survey](https://mlsanity.com/paper/2406.05424/) (2024)
- [Modeling Target-side Inflection in Placeholder Translation](https://mlsanity.com/paper/2107.00334/) (2021)
- [Modeling Target-Side Morphology in Neural Machine Translation: A Comparison of Strategies](https://mlsanity.com/paper/2203.13550/) (2022)
- [Understanding the effects of word-level linguistic annotations in under-resourced neural machine translation](https://mlsanity.com/paper/2401.16078/) (2024)

## Abstract (arXiv)

NMT systems have problems with large vocabulary sizes. Byte-pair encoding (BPE) is a popular approach to solving this problem, but while BPE allows the system to generate any target-side word, it does not enable effective generalization over the rich vocabulary in morphologically rich languages with strong inflectional phenomena. We introduce a simple approach to overcome this problem by training a system to produce the lemma of a word and its morphologically rich POS tag, which is then followed by a deterministic generation step. We apply this strategy for English-Czech and English-German translation scenarios, obtaining improvements in both settings. We furthermore show that the improvement is not due to only adding explicit morphological information.