LTNER: Large Language Model Tagging for Named Entity Recognition with Contextualized Entity Marking

TL;DR

This paper introduces LTNER, a framework that enhances large language models' performance on named entity recognition by using a novel contextualized entity marking method, achieving near-supervised accuracy without additional training.

Contribution

The paper presents a new method called Contextualized Entity Marking Gen that significantly improves LLMs' NER accuracy without extra training, bridging the gap with supervised methods.

Findings

F1 score on CoNLL03 increased from 85.9% to 91.9%.

Achieved near-supervised performance without additional training.

Demonstrated the potential of LLMs in NER tasks.

Abstract

The use of LLMs for natural language processing has become a popular trend in the past two years, driven by their formidable capacity for context comprehension and learning, which has inspired a wave of research from academics and industry professionals. However, for certain NLP tasks, such as NER, the performance of LLMs still falls short when compared to supervised learning methods. In our research, we developed a NER processing framework called LTNER that incorporates a revolutionary Contextualized Entity Marking Gen Method. By leveraging the cost-effective GPT-3.5 coupled with context learning that does not require additional training, we significantly improved the accuracy of LLMs in handling NER tasks. The F1 score on the CoNLL03 dataset increased from the initial 85.9% to 91.9%, approaching the performance of supervised fine-tuning. This outcome has led to a deeper understanding of the potential of LLMs.