Creoles represent an under-explored and marginalized group of languages, with few available resources for NLP research.While the genealogical ties between Creoles and a number of highly-resourced languages imply a significant potential for transfer learning, this potential is hampered due to this lack of annotated data. In this work we present CreoleVal, a collection of benchmark datasets spanning 8 different NLP tasks, covering up to 28 Creole languages; it is an aggregate of novel development datasets for reading comprehension, relation classification, and machine translation for Creoles, in addition to a practical gateway to a handful of preexisting benchmarks. For each benchmark, we conduct baseline experiments in a zero-shot setting in order to further ascertain the capabilities and limitations of transfer learning for Creoles. Ultimately, we see CreoleVal as an opportunity to empower research on Creoles in NLP and computational linguistics, and in general, a step towards more equitable language technology around the globe.
@article{arxiv.2310.19567,
title = {CreoleVal: Multilingual Multitask Benchmarks for Creoles},
author = {Heather Lent and Kushal Tatariya and Raj Dabre and Yiyi Chen and Marcell Fekete and Esther Ploeger and Li Zhou and Ruth-Ann Armstrong and Abee Eijansantos and Catriona Malau and Hans Erik Heje and Ernests Lavrinovics and Diptesh Kanojia and Paul Belony and Marcel Bollmann and Loïc Grobol and Miryam de Lhoneux and Daniel Hershcovich and Michel DeGraff and Anders Søgaard and Johannes Bjerva},
journal= {arXiv preprint arXiv:2310.19567},
year = {2024}
}