Enhancing Named Entity Recognition in Low Resource Languages through Few-Shot Learning

Abstract

This research addresses the challenge of developing effective Named Entity Recognition (NER) models for low-resource languages, with a specific focus on Tamil. While demand for NER systems is growing, low-resource languages often lack the extensive labeled data required to train high-performance models. To bridge this gap, we propose a novel approach utilizing meta-learning and prototypical networks to significantly improve NER accuracy in Tamil text.