Pivot Language for Low-Resource Machine Translation

TL;DR

This paper explores using Hindi as a pivot language to improve Nepali-English translation, employing transfer and backtranslation methods, achieving notable BLEU score improvements over previous baselines.

Contribution

It demonstrates the effectiveness of Hindi as a pivot language for Nepali-English translation and compares supervised and semi-supervised approaches, highlighting areas for future enhancement.

Findings

Transfer method achieves BLEU score of 14.2, surpassing previous baseline by 6.6 points.

Backtranslation approach scores 15.1 BLEU, slightly below semi-supervised baseline.

Analysis discusses factors affecting performance and future research directions.

Abstract

Certain pairs of languages suffer from lack of a parallel corpus which is large in size and diverse in domain. One of the ways this is overcome is via use of a pivot language. In this paper we use Hindi as a pivot language to translate Nepali into English. We describe what makes Hindi a good candidate for the pivot. We discuss ways in which a pivot language can be used, and use two such approaches - the Transfer Method (fully supervised) and Backtranslation (semi-supervised) - to translate Nepali into English. Using the former, we are able to achieve a devtest Set SacreBLEU score of 14.2, which improves the baseline fully supervised score reported by (Guzman et al., 2019) by 6.6 points. While we are slightly below the semi-supervised baseline score of 15.1, we discuss what may have caused this under-performance, and suggest scope for future work.