High-quality Data-to-Text Generation for Severely Under-Resourced Languages with Out-of-the-box Large Language Models
This addresses the problem of limited NLP capabilities for under-resourced languages, showing incremental improvement by applying existing LLMs to new data.
The study tackled the performance gap in NLP for severely under-resourced languages like Irish, Welsh, Breton, and Maltese by testing pretrained large language models (LLMs) for data-to-text generation, finding that LLMs set new state-of-the-art results with human-evaluated performance on par with humans.
The performance of NLP methods for severely under-resourced languages cannot currently hope to match the state of the art in NLP methods for well resourced languages. We explore the extent to which pretrained large language models (LLMs) can bridge this gap, via the example of data-to-text generation for Irish, Welsh, Breton and Maltese. We test LLMs on these under-resourced languages and English, in a range of scenarios. We find that LLMs easily set the state of the art for the under-resourced languages by substantial margins, as measured by both automatic and human evaluations. For all our languages, human evaluation shows on-a-par performance with humans for our best systems, but BLEU scores collapse compared to English, casting doubt on the metric's suitability for evaluating non-task-specific systems. Overall, our results demonstrate the great potential of LLMs to bridge the performance gap for under-resourced languages.