EntGPT: Entity Linking with Generative Large Language Models

TL;DR

EntGPT introduces advanced prompt engineering techniques for entity linking using large language models, significantly improving performance across multiple datasets without the need for supervised fine-tuning.

Contribution

The paper presents novel prompt engineering methods, including hard prompting and instruction tuning, to enhance entity linking performance with large language models.

Findings

Up to 36% improvement in micro-F_1 score with EntGPT-P

Average 2.1% increase in micro-F_1 score with EntGPT-I in supervised tasks

Outperforms baseline models in six Question Answering datasets

Abstract

Entity Linking in natural language processing seeks to match text entities to their corresponding entries in a dictionary or knowledge base. Traditional approaches rely on contextual models, which can be complex, hard to train, and have limited transferability across different domains. Generative large language models like GPT offer a promising alternative but often underperform with naive prompts. In this study, we introduce EntGPT, employing advanced prompt engineering to enhance EL tasks. Our three-step hard-prompting method (EntGPT-P) significantly boosts the micro-F_1 score by up to 36% over vanilla prompts, achieving competitive performance across 10 datasets without supervised fine-tuning. Additionally, our instruction tuning method (EntGPT-I) improves micro-F_1 scores by 2.1% on average in supervised EL tasks and outperforms several baseline models in six Question Answering tasks. Our methods are compatible with both open-source and proprietary LLMs. All data and code are available on GitHub at https://github.com/yifding/In_Context_EL.