PRiSM: Enhancing Low-Resource Document-Level Relation Extraction with Relation-Aware Score Calibration

Minseok Choi, Hyesu Lim, Jaegul Choo
9/25/2023
cs.CLcs.AIcs.LG

Official code: brightjade/PRiSM (11 stars, MIT, last commit 2025-06-20).

Abstract

Document-level relation extraction (DocRE) aims to extract relations of all entity pairs in a document. A key challenge in DocRE is the cost of annotating such data which requires intensive human effort. Thus, we investigate the case of DocRE in a low-resource setting, and we find that existing models trained on low data overestimate the NA ("no relation") label, causing limited performance. In this work, we approach the problem from a calibration perspective and propose PRiSM, which learns to adapt logits based on relation semantic information. We evaluate our method on three DocRE datasets and demonstrate that integrating existing models with PRiSM improves performance by as much as 26.38 F1 score, while the calibration error drops as much as 36 times when trained with about 3% of data. The code is publicly available at https://github.com/brightjade/PRiSM.

View on arXivView PDF

Code Implementations(1)

brightjade/PRiSMOfficial100%
MIT

Source code for paper "PRiSM: Enhancing Low-Resource Document-Level Relation Extraction with Relation-Aware Score Calibration", Findings of IJCNLP-AACL 2023

110Shell, PythonSep 14, 20231 years agoMIT
calibrationdocument-leveldocument-level-relation-extractionlow-resourcerelation-extraction

Similar Papers

DREEAM: Guiding Attention with Evidence for Improving Document-Level Relation Extraction

86% match1 repo942023

Class-Adaptive Self-Training for Relation Extraction with Incompletely Annotated Training Data

86% match1 repo52023

Retrieval-Augmented Generation-based Relation Extraction

86% match1 repo492024

Siamese Representation Learning for Unsupervised Relation Extraction

85% match1 repo42023

Enhancing Low-Resource Relation Representations through Multi-View Decoupling

85% match1 repo102023

Sentence-Level Relation Extraction via Contrastive Learning with Descriptive Relation Prompts

85% match2023

Cite this paper

@article{choi2023prism,
  title  = {PRiSM: Enhancing Low-Resource Document-Level Relation Extraction with Relation-Aware Score Calibration},
  author = {Minseok Choi and Hyesu Lim and Jaegul Choo},
  year   = {2023},
  eprint = {2309.13869},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2309.13869v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2309.13869.svg)](https://codeofpaper.com/paper/2309.13869)

Discussion