A Prototypical Semantic Decoupling Method via Joint Contrastive Learning for Few-Shot Name Entity Recognition

TL;DR

This paper introduces PSDC, a novel few-shot NER method that decouples semantic information via contrastive learning, leading to improved performance and better generalization over previous state-of-the-art models.

Contribution

The paper proposes a new decoupling approach with joint contrastive learning for few-shot NER, addressing prototype confusion and enhancing semantic representation.

Findings

Outperforms previous SOTA methods on two benchmarks

Effectively decouples class-specific and contextual semantics

Demonstrates strong generalization and robustness

Abstract

Few-shot named entity recognition (NER) aims at identifying named entities based on only few labeled instances. Most existing prototype-based sequence labeling models tend to memorize entity mentions which would be easily confused by close prototypes. In this paper, we proposed a Prototypical Semantic Decoupling method via joint Contrastive learning (PSDC) for few-shot NER. Specifically, we decouple class-specific prototypes and contextual semantic prototypes by two masking strategies to lead the model to focus on two different semantic information for inference. Besides, we further introduce joint contrastive learning objectives to better integrate two kinds of decoupling information and prevent semantic collapse. Experimental results on two few-shot NER benchmarks demonstrate that PSDC consistently outperforms the previous SOTA methods in terms of overall performance. Extensive analysis further validates the effectiveness and generalization of PSDC.