Few-shot Named Entity Recognition with Self-describing Networks

TL;DR

This paper introduces Self-describing Networks (SDNet), a novel approach for few-shot NER that leverages a universal concept set to improve entity recognition with limited data, achieving state-of-the-art results.

Contribution

The paper proposes a self-describing mechanism and SDNet model that effectively transfer knowledge and recognize entities in few-shot NER by describing mentions with concepts.

Findings

SDNet achieves new state-of-the-art on 6 out of 8 benchmarks.

The model effectively leverages external resources and a universal concept set.

Experiments demonstrate robustness across diverse domains.

Abstract

Few-shot NER needs to effectively capture information from limited instances and transfer useful knowledge from external resources. In this paper, we propose a self-describing mechanism for few-shot NER, which can effectively leverage illustrative instances and precisely transfer knowledge from external resources by describing both entity types and mentions using a universal concept set. Specifically, we design Self-describing Networks (SDNet), a Seq2Seq generation model which can universally describe mentions using concepts, automatically map novel entity types to concepts, and adaptively recognize entities on-demand. We pre-train SDNet with large-scale corpus, and conduct experiments on 8 benchmarks from different domains. Experiments show that SDNet achieves competitive performances on all benchmarks and achieves the new state-of-the-art on 6 benchmarks, which demonstrates its effectiveness and robustness.