Alligators All Around: Mitigating Lexical Confusion in Low-resource Machine Translation

Abstract

Current machine translation (MT) systems for low-resource languages have a particular failure mode: When translating words in a given domain, they tend to confuse words within that domain. So, for example, lion might be translated as alligator, and orange might be rendered as purple. We propose a recall-based metric for measuring this problem and show that the problem exists in a dataset comprising 122 low-resource languages. We then show that this problem can be mitigated by using a large language model (LLM) to post-edit the MT output, specifically by including the entire GATITOS lexicon for the relevant language as a very long context prompt. We show gains in average CHRF score over the set of 122 languages, and we show that the recall score for relevant lexical items also improves. Finally, we demonstrate that a small dedicated MT system with a general-purpose LLM as a post-editor outperforms a generalist LLM translator with access to the same lexicon data, suggesting a new paradigm for LLM use.