计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (9): 1-12.DOI: 10.3778/j.issn.1002-8331.2208-0027

• 热点与综述 • 上一篇    下一篇

少样本关系分类综述

刘涛,柯尊旺,吾守尔·斯拉木   

  1. 1.新疆大学 软件学院,乌鲁木齐 830046
    2.新疆多语种信息技术实验室,乌鲁木齐 830046
    3.新疆多语种信息技术研究中心,乌鲁木齐 830046
  • 出版日期:2023-05-01 发布日期:2023-05-01

Survey of Few-Shot Relation Classification

LIU Tao, KE Zunwang, Wushour·Silamu   

  1. 1.School of Software, Xinjiang University, Urumqi 830046, China
    2.Xinjiang Multilingual Information Technology Laboratory, Urumqi 830046, China
    3.Xinjiang Multilingual Information Technology Research Center, Urumqi 830046, China
  • Online:2023-05-01 Published:2023-05-01

摘要: 少样本关系分类旨在通过少量的有标注训练样本,来挖掘自然语言文本中目标实体之间所蕴含的语义关系,以应对传统的关系分类方法所面临的资源匮乏问题,从而能够较好地推广到医学、金融以及民语处理等数据稀缺的特定领域。目前,少样本关系分类的相关研究工作均在元学习的训练策略下学习先验知识,并以此快速适应新的任务,其大体上可以划分为基于原型网络、基于预训练语言模型、基于参数优化以及基于图神经网络四种方式。回顾少样本关系分类的发展,对不同研究方法的优势和局限性进行深入剖析和总结,在此基础上,分析该领域当前所面临的棘手问题和挑战,并进一步对其未来的研究方向进行展望。

关键词: 关系分类, 少样本学习, 度量学习, 元学习

Abstract: Few-shot relation classification aims to mine the semantic relationship between target entities in natural language texts with limited labeled training examples, so as to deal with the resource shortage problem faced by the traditional relation classification methods, so that it can be better applied to medicine, finance and ethnic language processing and other data scarce fields. At present, the relevant research work on few-shot relation classification all learns prior knowledge under the training strategy of meta learning, and to quickly adapt to new tasks. Generally, it can be divided into four classes method:prototype network based, pre-training language model based, parameter optimization based, and graph neural network based. This paper reviews the development of few-shot relation classification, analyzes and summarizes the advantages and limitations of different research methods. On this basis, the paper analyzes the current problems and challenge faced by few-shot relation classification, and prospects the future research directions of this field.

Key words: relation classification, few-shot learning, metric learning, meta learning