结合元学习和提示学习的Few-Shot关系抽取研究

IF 6.5 2区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE
Tingting Hang , Wei Wu , Jun Feng , Hamza Djigal , Jun Huang
{"title":"结合元学习和提示学习的Few-Shot关系抽取研究","authors":"Tingting Hang ,&nbsp;Wei Wu ,&nbsp;Jun Feng ,&nbsp;Hamza Djigal ,&nbsp;Jun Huang","doi":"10.1016/j.neucom.2025.130534","DOIUrl":null,"url":null,"abstract":"<div><div>Few-shot Relation Extraction (FSRE) is challenging in Natural Language Processing (NLP). It requires models to accurately identify and categorize the relation between entities in texts based on limited annotated samples. Researchers have explored meta-learning and prompt-learning methods to address this challenge. Meta-learning aims to accelerate the model’s learning rate for new tasks by training across different tasks, while prompt learning seeks to enhance the model’s understanding of context. However, both methods face limitations in few-shot scenarios. To overcome these restrictions, studies combining meta-learning with prompt learning have emerged, garnering widespread academic interest. This paper presents a comprehensive survey of FSRE combining Meta-Learning with Prompt Learning. We first categorize existing research into three main groups: meta-learning-based, prompt learning-based, and hybrid approaches. Then, we provide a detailed analysis and discussion of the latest advancements in each category. Moreover, we extensively describe datasets commonly used in the FSRE task, evaluation metrics, and comparisons among various models. Finally, we explore current challenges in the field and forecast potential future research trends, offering insights for subsequent studies.</div></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":"647 ","pages":"Article 130534"},"PeriodicalIF":6.5000,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A survey of Few-Shot Relation Extraction combining meta-learning with prompt learning\",\"authors\":\"Tingting Hang ,&nbsp;Wei Wu ,&nbsp;Jun Feng ,&nbsp;Hamza Djigal ,&nbsp;Jun Huang\",\"doi\":\"10.1016/j.neucom.2025.130534\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Few-shot Relation Extraction (FSRE) is challenging in Natural Language Processing (NLP). It requires models to accurately identify and categorize the relation between entities in texts based on limited annotated samples. Researchers have explored meta-learning and prompt-learning methods to address this challenge. Meta-learning aims to accelerate the model’s learning rate for new tasks by training across different tasks, while prompt learning seeks to enhance the model’s understanding of context. However, both methods face limitations in few-shot scenarios. To overcome these restrictions, studies combining meta-learning with prompt learning have emerged, garnering widespread academic interest. This paper presents a comprehensive survey of FSRE combining Meta-Learning with Prompt Learning. We first categorize existing research into three main groups: meta-learning-based, prompt learning-based, and hybrid approaches. Then, we provide a detailed analysis and discussion of the latest advancements in each category. Moreover, we extensively describe datasets commonly used in the FSRE task, evaluation metrics, and comparisons among various models. Finally, we explore current challenges in the field and forecast potential future research trends, offering insights for subsequent studies.</div></div>\",\"PeriodicalId\":19268,\"journal\":{\"name\":\"Neurocomputing\",\"volume\":\"647 \",\"pages\":\"Article 130534\"},\"PeriodicalIF\":6.5000,\"publicationDate\":\"2025-06-02\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Neurocomputing\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0925231225012068\",\"RegionNum\":2,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Neurocomputing","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0925231225012068","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0

摘要

在自然语言处理(NLP)中,少镜头关系提取(FSRE)是一个具有挑战性的问题。它要求模型基于有限的注释样本,准确地识别和分类文本中实体之间的关系。研究人员已经探索了元学习和即时学习方法来应对这一挑战。元学习旨在通过跨任务训练来加快模型对新任务的学习率,而提示学习旨在增强模型对上下文的理解。然而,这两种方法在少量射击场景中都有局限性。为了克服这些限制,将元学习与即时学习结合起来的研究已经出现,并引起了广泛的学术兴趣。本文对元学习与提示学习相结合的FSRE进行了综述。我们首先将现有的研究分为三大类:基于元学习的方法、基于提示学习的方法和混合方法。然后,我们对每个类别的最新进展进行了详细的分析和讨论。此外,我们还广泛描述了FSRE任务中常用的数据集、评估指标以及各种模型之间的比较。最后,我们探讨了该领域当前面临的挑战,并预测了潜在的未来研究趋势,为后续研究提供了见解。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
A survey of Few-Shot Relation Extraction combining meta-learning with prompt learning
Few-shot Relation Extraction (FSRE) is challenging in Natural Language Processing (NLP). It requires models to accurately identify and categorize the relation between entities in texts based on limited annotated samples. Researchers have explored meta-learning and prompt-learning methods to address this challenge. Meta-learning aims to accelerate the model’s learning rate for new tasks by training across different tasks, while prompt learning seeks to enhance the model’s understanding of context. However, both methods face limitations in few-shot scenarios. To overcome these restrictions, studies combining meta-learning with prompt learning have emerged, garnering widespread academic interest. This paper presents a comprehensive survey of FSRE combining Meta-Learning with Prompt Learning. We first categorize existing research into three main groups: meta-learning-based, prompt learning-based, and hybrid approaches. Then, we provide a detailed analysis and discussion of the latest advancements in each category. Moreover, we extensively describe datasets commonly used in the FSRE task, evaluation metrics, and comparisons among various models. Finally, we explore current challenges in the field and forecast potential future research trends, offering insights for subsequent studies.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Neurocomputing
Neurocomputing 工程技术-计算机:人工智能
CiteScore
13.10
自引率
10.00%
发文量
1382
审稿时长
70 days
期刊介绍: Neurocomputing publishes articles describing recent fundamental contributions in the field of neurocomputing. Neurocomputing theory, practice and applications are the essential topics being covered.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信