Does Synthetic Data Help Named Entity Recognition for Low-Resource Languages?

Gaurav Kamath, Sowmya Vajjala
5/22/2025
cs.CL

No official code found. Best implementation: grvkamath/low-resource-syn-ner (0 stars, MIT, last commit 2025-11-05).

Abstract

Named Entity Recognition(NER) for low-resource languages aims to produce robust systems for languages where there is limited labeled training data available, and has been an area of increasing interest within NLP. Data augmentation for increasing the amount of low-resource labeled data is a common practice. In this paper, we explore the role of synthetic data in the context of multilingual, low-resource NER, considering 11 languages from diverse language families. Our results suggest that synthetic data does in fact hold promise for low-resource language NER, though we see significant variation between languages.

View on arXivView PDF

Code Implementations(1)

grvkamath/low-resource-syn-nerCommunity67%
MIT

Code and data used in the paper "Does Synthetic Data Help Named Entity Recognition for Low-Resource Languages?", to be presented at AACL 2025 (Main Conference).

00R, PythonNov 5, 202511 months agoMIT

Cite this paper

@article{kamath2025does,
  title  = {Does Synthetic Data Help Named Entity Recognition for Low-Resource Languages?},
  author = {Gaurav Kamath and Sowmya Vajjala},
  year   = {2025},
  eprint = {2505.16814},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2505.16814v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2505.16814.svg)](https://codeofpaper.com/paper/2505.16814)

Discussion