The paper discusses the use of Large Language Models (LLMs) to improve the performance of BioMedical Named Entity Recognition (NER). Traditional methods often struggle with NER tasks, particularly those requiring domain-specific knowledge. The authors propose a two-step process, first extracting entity boundaries with the LLM, and then determining the entity category. This process is enhanced by injecting external knowledge to assist with category determination. The paper claims this approach significantly improves BioNER performance compared to previous methods.
Publication date: 21 Sep 2023
Project Page: https://arxiv.org/abs/2309.12278v1
Paper: https://arxiv.org/pdf/2309.12278
