Enregistré dans:
Détails bibliographiques
Auteurs principaux: Han, Peitao, Pereira, Lis Kanashiro, Cheng, Fei, She, Wan Jou, Aramaki, Eiji
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2406.10432
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909592791810048
author Han, Peitao
Pereira, Lis Kanashiro
Cheng, Fei
She, Wan Jou
Aramaki, Eiji
author_facet Han, Peitao
Pereira, Lis Kanashiro
Cheng, Fei
She, Wan Jou
Aramaki, Eiji
contents Existing in-context learning (ICL) methods for relation extraction (RE) often prioritize language similarity over structural similarity, which can lead to overlooking entity relationships. To address this, we propose an AMR-enhanced retrieval-based ICL method for RE. Our model retrieves in-context examples based on semantic structure similarity between task inputs and training samples. Evaluations on four standard English RE datasets show that our model outperforms baselines in the unsupervised setting across all datasets. In the supervised setting, it achieves state-of-the-art results on three datasets and competitive results on the fourth.
format Preprint
id arxiv_https___arxiv_org_abs_2406_10432
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle AMR-RE: Abstract Meaning Representations for Retrieval-Based In-Context Learning in Relation Extraction
Han, Peitao
Pereira, Lis Kanashiro
Cheng, Fei
She, Wan Jou
Aramaki, Eiji
Computation and Language
Existing in-context learning (ICL) methods for relation extraction (RE) often prioritize language similarity over structural similarity, which can lead to overlooking entity relationships. To address this, we propose an AMR-enhanced retrieval-based ICL method for RE. Our model retrieves in-context examples based on semantic structure similarity between task inputs and training samples. Evaluations on four standard English RE datasets show that our model outperforms baselines in the unsupervised setting across all datasets. In the supervised setting, it achieves state-of-the-art results on three datasets and competitive results on the fourth.
title AMR-RE: Abstract Meaning Representations for Retrieval-Based In-Context Learning in Relation Extraction
topic Computation and Language
url https://arxiv.org/abs/2406.10432