Robust Self-Augmentation for Named Entity Recognition with Meta Reweighting

TL;DR

This paper introduces a unified meta-reweighting strategy to enhance self-augmentation methods like token substitution and mixup for improving NER performance, especially in low-resource settings.

Contribution

It proposes a flexible meta-reweighting approach that effectively integrates different self-augmentation techniques for NER without extensive customization.

Findings

Improved NER performance on Chinese and English benchmarks.

Meta-reweighting enhances the benefits of self-augmentation methods.

The approach is easily extensible to various self-augmentation techniques.

Abstract

Self-augmentation has received increasing research interest recently to improve named entity recognition (NER) performance in low-resource scenarios. Token substitution and mixup are two feasible heterogeneous self-augmentation techniques for NER that can achieve effective performance with certain specialized efforts. Noticeably, self-augmentation may introduce potentially noisy augmented data. Prior research has mainly resorted to heuristic rule-based constraints to reduce the noise for specific self-augmentation methods individually. In this paper, we revisit these two typical self-augmentation methods for NER, and propose a unified meta-reweighting strategy for them to achieve a natural integration. Our method is easily extensible, imposing little effort on a specific self-augmentation method. Experiments on different Chinese and English NER benchmarks show that our token substitution and mixup method, as well as their integration, can achieve effective performance improvement. Based on the meta-reweighting mechanism, we can enhance the advantages of the self-augmentation techniques without much extra effort.