Enregistré dans:
Détails bibliographiques
Auteurs principaux: Yoo, Jeongkyun, Riddle, Nela, Hoblitzell, Andrew
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2508.16833
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866916913394745344
author Yoo, Jeongkyun
Riddle, Nela
Hoblitzell, Andrew
author_facet Yoo, Jeongkyun
Riddle, Nela
Hoblitzell, Andrew
contents Named Entity Recognition (NER) in biomedical domains faces challenges due to data scarcity and imbalanced label distributions, especially with fine-grained entity types. We propose ReProCon, a novel few-shot NER framework that combines multi-prototype modeling, cosine-contrastive learning, and Reptile meta-learning to tackle these issues. By representing each category with multiple prototypes, ReProCon captures semantic variability, such as synonyms and contextual differences, while a cosine-contrastive objective ensures strong interclass separation. Reptile meta-updates enable quick adaptation with little data. Using a lightweight fastText + BiLSTM encoder with much lower memory usage, ReProCon achieves a macro-$F_1$ score close to BERT-based baselines (around 99 percent of BERT performance). The model remains stable with a label budget of 30 percent and only drops 7.8 percent in $F_1$ when expanding from 19 to 50 categories, outperforming baselines such as SpanProto and CONTaiNER, which see 10 to 32 percent degradation in Few-NERD. Ablation studies highlight the importance of multi-prototype modeling and contrastive learning in managing class imbalance. Despite difficulties with label ambiguity, ReProCon demonstrates state-of-the-art performance in resource-limited settings, making it suitable for biomedical applications.
format Preprint
id arxiv_https___arxiv_org_abs_2508_16833
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ReProCon: Scalable and Resource-Efficient Few-Shot Biomedical Named Entity Recognition
Yoo, Jeongkyun
Riddle, Nela
Hoblitzell, Andrew
Computation and Language
Named Entity Recognition (NER) in biomedical domains faces challenges due to data scarcity and imbalanced label distributions, especially with fine-grained entity types. We propose ReProCon, a novel few-shot NER framework that combines multi-prototype modeling, cosine-contrastive learning, and Reptile meta-learning to tackle these issues. By representing each category with multiple prototypes, ReProCon captures semantic variability, such as synonyms and contextual differences, while a cosine-contrastive objective ensures strong interclass separation. Reptile meta-updates enable quick adaptation with little data. Using a lightweight fastText + BiLSTM encoder with much lower memory usage, ReProCon achieves a macro-$F_1$ score close to BERT-based baselines (around 99 percent of BERT performance). The model remains stable with a label budget of 30 percent and only drops 7.8 percent in $F_1$ when expanding from 19 to 50 categories, outperforming baselines such as SpanProto and CONTaiNER, which see 10 to 32 percent degradation in Few-NERD. Ablation studies highlight the importance of multi-prototype modeling and contrastive learning in managing class imbalance. Despite difficulties with label ambiguity, ReProCon demonstrates state-of-the-art performance in resource-limited settings, making it suitable for biomedical applications.
title ReProCon: Scalable and Resource-Efficient Few-Shot Biomedical Named Entity Recognition
topic Computation and Language
url https://arxiv.org/abs/2508.16833