Effective Multi-Task Learning for Biomedical Named Entity Recognition

TL;DR

This paper presents SRU-NER, a multi-task learning model for biomedical NER that effectively handles nested entities and dataset inconsistencies, improving cross-domain generalization and performance.

Contribution

Introduction of SRU-NER, a novel multi-task learning approach that manages nested entities and dataset annotation gaps in biomedical NER tasks.

Findings

Achieves competitive performance in biomedical NER tasks

Improves cross-domain generalization

Effectively handles nested entities and dataset inconsistencies

Abstract

Biomedical Named Entity Recognition presents significant challenges due to the complexity of biomedical terminology and inconsistencies in annotation across datasets. This paper introduces SRU-NER (Slot-based Recurrent Unit NER), a novel approach designed to handle nested named entities while integrating multiple datasets through an effective multi-task learning strategy. SRU-NER mitigates annotation gaps by dynamically adjusting loss computation to avoid penalizing predictions of entity types absent in a given dataset. Through extensive experiments, including a cross-corpus evaluation and human assessment of the model's predictions, SRU-NER achieves competitive performance in biomedical and general-domain NER tasks, while improving cross-domain generalization.