The Effect of Domain and Diacritics in Yor\`ub\'a-English Neural Machine Translation

TL;DR

This paper introduces MENYO-20k, a standardized Yorùbá-English dataset, and evaluates neural machine translation models, highlighting the impact of diacritics on translation quality and outperforming existing multilingual models.

Contribution

The creation of MENYO-20k dataset and comprehensive benchmarks for Yorùbá-English translation, including analysis of diacritics' effects on translation quality.

Findings

Models outperform Google and Facebook M2M in Yorùbá translation.

Diacritics significantly influence translation quality and intelligibility.

Benchmark results set a new standard for Yorùbá-English neural machine translation.

Abstract

Massively multilingual machine translation (MT) has shown impressive capabilities, including zero and few-shot translation between low-resource language pairs. However, these models are often evaluated on high-resource languages with the assumption that they generalize to low-resource ones. The difficulty of evaluating MT models on low-resource pairs is often due to lack of standardized evaluation datasets. In this paper, we present MENYO-20k, the first multi-domain parallel corpus with a special focus on clean orthography for Yor\`ub\'a--English with standardized train-test splits for benchmarking. We provide several neural MT benchmarks and compare them to the performance of popular pre-trained (massively multilingual) MT models both for the heterogeneous test set and its subdomains. Since these pre-trained models use huge amounts of data with uncertain quality, we also analyze the effect of diacritics, a major characteristic of Yor\`ub\'a, in the training data. We investigate how and when this training condition affects the final quality and intelligibility of a translation. Our models outperform massively multilingual models such as Google ($+8.7$ BLEU) and Facebook M2M ($+9.1$ BLEU) when translating to Yor\`ub\'a, setting a high quality benchmark for future research.