Do "English" Named Entity Recognizers Work Well on Global Englishes?

TL;DR

This study evaluates how well existing English NER models perform on diverse global English varieties, revealing significant performance drops on non-American/British English data and suggesting the need for more inclusive training datasets.

Contribution

The paper introduces the Worldwide English NER Dataset and analyzes NER model performance across diverse English varieties, highlighting limitations of current models on global English data.

Findings

Significant performance drops (>10 F1) on global English varieties.

Models trained on standard datasets perform poorly on Oceania and Africa.

Combined training on the new dataset improves robustness with minimal F1 loss.

Abstract

The vast majority of the popular English named entity recognition (NER) datasets contain American or British English data, despite the existence of many global varieties of English. As such, it is unclear whether they generalize for analyzing use of English globally. To test this, we build a newswire dataset, the Worldwide English NER Dataset, to analyze NER model performance on low-resource English variants from around the world. We test widely used NER toolkits and transformer models, including models using the pre-trained contextual models RoBERTa and ELECTRA, on three datasets: a commonly used British English newswire dataset, CoNLL 2003, a more American focused dataset OntoNotes, and our global dataset. All models trained on the CoNLL or OntoNotes datasets experienced significant performance drops-over 10 F1 in some cases-when tested on the Worldwide English dataset. Upon examination of region-specific errors, we observe the greatest performance drops for Oceania and Africa, while Asia and the Middle East had comparatively strong performance. Lastly, we find that a combined model trained on the Worldwide dataset and either CoNLL or OntoNotes lost only 1-2 F1 on both test sets.