Can we teach language models to gloss endangered languages?

TL;DR

This paper investigates the use of large language models with in-context learning to automate interlinear glossed text creation for endangered languages, showing promising results without traditional training.

Contribution

It introduces novel example selection strategies for LLMs in glossing tasks and demonstrates their effectiveness compared to standard baselines.

Findings

LLMs outperform standard transformer baselines in glossing tasks

Targeted example selection significantly improves LLM performance

LLMs require no traditional training, making them practical for language documentation

Abstract

Interlinear glossed text (IGT) is a popular format in language documentation projects, where each morpheme is labeled with a descriptive annotation. Automating the creation of interlinear glossed text would be desirable to reduce annotator effort and maintain consistency across annotated corpora. Prior research has explored a number of statistical and neural methods for automatically producing IGT. As large language models (LLMs) have showed promising results across multilingual tasks, even for rare, endangered languages, it is natural to wonder whether they can be utilized for the task of generating IGT. We explore whether LLMs can be effective at the task of interlinear glossing with in-context learning, without any traditional training. We propose new approaches for selecting examples to provide in-context, observing that targeted selection can significantly improve performance. We find that LLM-based methods beat standard transformer baselines, despite requiring no training at all. These approaches still underperform state-of-the-art supervised systems for the task, but are highly practical for researchers outside of the NLP community, requiring minimal effort to use.