LLM-Augmented Prototype Representation for Few-Shot Named-Entity Recognition
Open Access
- 11 November 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 198792-198811
Abstract
Named Entity Recognition (NER) models face challenges in adapting to data distribution shifts, especially with unseen entity types and limited data. Few-shot learning is used to address long-tailed distributions and unseen classes, but struggles with few labeled examples, leading to inaccurate prototypes. Recently, Large Language Models (LLMs) excel at Few-Shot Named Entity Recognition (FS-NER) via in-context learning, but their substantially higher computational cost and latency compared to prototype-based models limit their practicality for resource-constrained or real-time applications. In this paper, we propose a new way to use LLMs to augment the FS-NER system and a way for the prototype network to utilize the augmented dataset. We aim to examine whether LLM In-context learning (LLM-ICL) can effectively replace the encoder model in the FS-NER task. We study an LLM-based data augmentation framework that generates new mentions and descriptions for specific entity types when fine-tuning, while maintaining inference-time efficiency, improving the robustness of embeddings and prototype representations. Our method shows consistent improvements across FS-NER setups, demonstrating the potential of LLMs as external knowledge sources for NER.Keywords
This publication has 17 references indexed in Scilit:
- ANeTCM: A Novel MRC Framework for Traditional Chinese Medicine Named Entity RecognitionIEEE Access, 2024
- Towards Pointsets Representation Learning via Self-Supervised Learning and Set AugmentationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2021
- Cluster Quality Analysis Using Silhouette Score2020
- Domain Generalization for Named Entity Boundary Detection via MetalearningIEEE Transactions on Neural Networks and Learning Systems, 2020
- Named Entity Extraction for Knowledge Graphs: A Literature OverviewIEEE Access, 2020
- Fine-grained Type Inference in Knowledge Graphs via Probabilistic and Tensor Factorization Methods2019
- Few-shot classification in named entity recognition task2019
- An Optimal Transportation Approach for Assessing Almost Stochastic Order2018
- The GUM corpus: creating multilayer resources in the classroomLanguage Resources and Evaluation, 2016
- FaceNet: A unified embedding for face recognition and clustering2015