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