Conference paper Open Access

Multitask Learning For Different Subword Segmentations In Neural Machine Translation

Srinivasan, Tejas; Sanabria, Ramon; Metze, Florian

In Neural Machine Translation (NMT) the usage of sub􏰃words and characters as source and target units offers a simple and flexible solution for translation of rare and unseen words. However, selecting the optimal subword segmentation involves a trade-off between expressiveness and flexibility, and is language and dataset-dependent. We present Block Multitask Learning (BMTL), a novel NMT architecture that predicts multiple targets of different granularities simultaneously, removing the need to search for the optimal segmentation strategy. Our multi-task model exhibits improvements of up to 1.7 BLEU points on each decoder over single-task baseline models with the same number of parameters on datasets from two language pairs of IWSLT15 and one from IWSLT19. The multiple hypotheses generated at different granularities can be combined as a post-processing step to give better translations, which improves over hypothesis combination from baseline models while using substantially fewer parameters.

Files (847.4 kB)
Name Size
847.4 kB Download
All versions This version
Views 7273
Downloads 6767
Data volume 56.8 MB56.8 MB
Unique views 6364
Unique downloads 6262


Cite as