A Robust and Domain-Adaptive Approach for Low-Resource Named Entity Recognition

TL;DR

This paper introduces RDANER, a robust and domain-adaptive method for low-resource NER that relies solely on inexpensive, easily accessible resources, achieving competitive performance without domain-specific knowledge bases.

Contribution

The paper presents a novel low-resource NER approach that does not depend on costly domain-specific resources, broadening applicability and reducing implementation barriers.

Findings

Achieves top performance with cheap resources on benchmark datasets.

Outperforms existing methods relying on domain-specific resources.

Demonstrates robustness across multiple domains.

Abstract

Recently, it has attracted much attention to build reliable named entity recognition (NER) systems using limited annotated data. Nearly all existing works heavily rely on domain-specific resources, such as external lexicons and knowledge bases. However, such domain-specific resources are often not available, meanwhile it's difficult and expensive to construct the resources, which has become a key obstacle to wider adoption. To tackle the problem, in this work, we propose a novel robust and domain-adaptive approach RDANER for low-resource NER, which only uses cheap and easily obtainable resources. Extensive experiments on three benchmark datasets demonstrate that our approach achieves the best performance when only using cheap and easily obtainable resources, and delivers competitive results against state-of-the-art methods which use difficultly obtainable domainspecific resources. All our code and corpora can be found on https://github.com/houking-can/RDANER.