English

Guideline Learning for In-context Information Extraction

Computation and Language 2025-07-14 v2 Machine Learning

Abstract

Large language models (LLMs) can perform a new task by merely conditioning on task instructions and a few input-output examples, without optimizing any parameters. This is called In-Context Learning (ICL). In-context Information Extraction (IE) has recently garnered attention in the research community. However, the performance of In-context IE generally lags behind the state-of-the-art supervised expert models. We highlight a key reason for this shortfall: underspecified task description. The limited-length context struggles to thoroughly express the intricate IE task instructions and various edge cases, leading to misalignment in task comprehension with humans. In this paper, we propose a Guideline Learning (GL) framework for In-context IE which reflectively learns and follows guidelines. During the learning phrase, GL automatically synthesizes a set of guidelines based on a few error cases, and during inference, GL retrieves helpful guidelines for better ICL. Moreover, we propose a self-consistency-based active learning method to enhance the efficiency of GL. Experiments on event extraction and relation extraction show that GL can significantly improve the performance of in-context IE.

Keywords

Cite

@article{arxiv.2310.05066,
  title  = {Guideline Learning for In-context Information Extraction},
  author = {Chaoxu Pang and Yixuan Cao and Qiang Ding and Ping Luo},
  journal= {arXiv preprint arXiv:2310.05066},
  year   = {2025}
}

Comments

EMNLP 2023 main conference

R2 v1 2026-06-28T12:43:45.818Z