Inspire the Large Language Model by External Knowledge on BioMedical Named Entity Recognition

TL;DR

This paper enhances biomedical named entity recognition by guiding large language models with external domain knowledge and a step-by-step approach, significantly improving their accuracy in extracting and classifying entities.

Contribution

It introduces a novel two-step BioNER method that leverages external knowledge and Chain-of-thought prompting to improve LLM performance in biomedical information extraction.

Findings

Significant improvement over previous few-shot LLM baselines.

External knowledge injection boosts entity category determination.

Step-by-step approach enhances overall BioNER accuracy.

Abstract

Large language models (LLMs) have demonstrated dominating performance in many NLP tasks, especially on generative tasks. However, they often fall short in some information extraction tasks, particularly those requiring domain-specific knowledge, such as Biomedical Named Entity Recognition (NER). In this paper, inspired by Chain-of-thought, we leverage the LLM to solve the Biomedical NER step-by-step: break down the NER task into entity span extraction and entity type determination. Additionally, for entity type determination, we inject entity knowledge to address the problem that LLM's lack of domain knowledge when predicting entity category. Experimental results show a significant improvement in our two-step BioNER approach compared to previous few-shot LLM baseline. Additionally, the incorporation of external knowledge significantly enhances entity category determination performance.