有监督实体关系联合抽取方法研究综述

doi:10.3778/j.issn.1673-9418.2107114

计算机科学与探索 ›› 2022, Vol. 16 ›› Issue (4): 713-733.DOI: 10.3778/j.issn.1673-9418.2107114

有监督实体关系联合抽取方法研究综述

张少伟¹, 王鑫¹^,²^,⁺(), 陈子睿¹, 王林³, 徐大为³, 贾勇哲¹^,³

1.天津大学智能与计算学部,天津 300350
2.天津市认知计算与应用重点实验室,天津 300350
3.天津泰凡科技有限公司,天津 300457

收稿日期:2021-07-21 修回日期:2022-03-04 出版日期:2022-04-01 发布日期:2022-04-14
通讯作者: + E-mail: wangx@tju.edu.cn
作者简介:张少伟（1996—）,男,硕士研究生,主要研究方向为知识表示学习、知识图谱构建。
王鑫（1981—）,男,博士,教授,博士生导师,CCF杰出会员,主要研究方向为知识图谱数据管理、图数据库、大数据分布式处理。
陈子睿（1998—）,男,硕士研究生,CCF学生会员,主要研究方向为知识表示学习、知识图谱问答。
王林（1981—）,男,博士,CCF专业会员,主要研究方向为大数据应用、人工智能。
徐大为（1989—）,男,博士,CCF专业会员,主要研究方向为人工智能、自然语言处理。
贾勇哲（1987—）,男,博士,CCF专业会员,主要研究方向为人工智能、先进制造业。
基金资助:
科技创新2030“新一代人工智能”重大项目(2020AAA0108504);国家自然科学基金面上项目(61972275)

Survey of Supervised Joint Entity Relation Extraction Methods

ZHANG Shaowei¹, WANG Xin¹^,²^,⁺(), CHEN Zirui¹, WANG Lin³, XU Dawei³, JIA Yongzhe¹^,³

1. College of Intelligence and Computing, Tianjin University, Tianjin 300350, China
2. Tianjin Key Laboratory of Cognitive Computing and Application, Tianjin 300350, China
3. Tianjin TechFantasy Co., Ltd., Tianjin 300457, China

Received:2021-07-21 Revised:2022-03-04 Online:2022-04-01 Published:2022-04-14
About author:ZHANG Shaowei, born in 1996, M.S. candidate. His research interests include knowledge repre-sentation learning and knowledge graph construc-tion.
WANG Xin, born in 1981, Ph.D., professor, Ph.D. supervisor, distinguished member of CCF. His research interests include knowledge graphs, graph databases and big data distributed processing.
CHEN Zirui, born in 1998, M.S. candidate, student member of CCF. His research interests include knowledge representation learning and knowledge graph question answering.
WANG Lin, born in 1981, Ph.D., professional member of CCF. His research interests include big data application and artificial intelligence.
XU Dawei, born in 1989, Ph.D., professional member of CCF. His research interests include artificial intelligence and natural language pro-cessing.
JIA Yongzhe, born in 1987, Ph.D., professional member of CCF. His research interests include artificial intelligence and advanced manufa-cturing.
Supported by:
Science and Technology Innovation 2030 “New Generation Artificial Intelligence” Major Project(2020AAA0108504);General Project of National Natural Science Foundation of China(61972275)

摘要/Abstract

摘要：

实体关系联合抽取作为信息抽取领域的核心任务,能够从非结构化或半结构化的文本中自动识别实体、实体类型以及实体之间特定的关系类型,为知识图谱构建、智能问答和语义搜索等下游任务提供基础支持。传统的流水线方法将实体关系联合抽取分解成命名实体识别和关系抽取两个独立的子任务,由于两个子任务之间缺少交互,流水线方法存在误差传播等问题。近年来,实体关系联合抽取成为新的研究趋势,其可以建立统一的模型使得不同子任务彼此交互,进一步提升模型性能。对有监督实体关系联合抽取方法进行综述,根据抽取特征的不同方式,可将实体关系联合抽取分为基于特征工程的联合抽取和基于神经网络的联合抽取两种类型。首先,介绍基于特征工程的联合抽取,包括整数线性规划、卡片金字塔解析、概率图模型和结构化预测四种方法,这四种方法都需要采用相对复杂的特征工程方法。然后,介绍基于神经网络的联合抽取,这类方法可以自动抽取特征信息,已逐渐成为联合抽取的主流方法,其主要包括共享参数和联合解码两种类型。接着,介绍有监督实体关系联合抽取常用的七个数据集以及评价指标,并对不同的实体关系联合抽取方法进行了实验对比分析。最后,展望实体关系联合抽取的未来研究方向。

关键词: 联合抽取, 特征工程, 神经网络

Abstract:

As a core task of information extraction, joint entity relation extraction can automatically identify entities, the types of entities and the specific relation between entities from unstructured texts or semi-structured texts, which provides basic support for downstream tasks such as knowledge graph construction, intelligent question answering, semantic search, etc. The traditional pipeline method decomposes joint entity relation extraction into two indepen-dent subtasks, named entity recognition and relation extraction. Due to the lack of interaction between the two subtasks, there are some problems such as error propagation in pipeline method. Recently, joint entity relation extraction has become a new trend, since it can further improve the performance of the model by establishing a unified model and making different subtasks interact. The supervised joint entity relation extraction approaches are surveyed in this paper. According to different ways of extracting features, there are two categories of joint entity relation extraction approaches, i.e., joint extraction based on feature engineering and joint extraction based on neural network. Firstly, the joint extraction based on feature engineering is introduced, including integer linear program-ming, card pyramid parsing, probabilistic graphical model and structured prediction, all of these four methods need to adopt complex feature engineering methods. Secondly, the joint extraction based on neural network is presented, which can automatically extract the feature information, gradually becoming the mainstream methods of joint extraction. Parameter sharing methods and joint decoding methods are two kinds of joint extraction methods based on neural network. Thirdly, seven common datasets and evaluation metrics of the supervised joint entity relation extraction are described, and the experimental comparison and analysis of different joint entity relation extraction methods are conducted. Finally, the future research directions of the joint entity relation extraction are put forward.

Key words: joint extraction, feature engineering, neural network

中图分类号:

TP311

张少伟, 王鑫, 陈子睿, 王林, 徐大为, 贾勇哲. 有监督实体关系联合抽取方法研究综述[J]. 计算机科学与探索, 2022, 16(4): 713-733.

ZHANG Shaowei, WANG Xin, CHEN Zirui, WANG Lin, XU Dawei, JIA Yongzhe. Survey of Supervised Joint Entity Relation Extraction Methods[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(4): 713-733.

图/表 22

图1 联合抽取方法分类

Fig.1 Classification of joint extraction methods

表1 常用符号描述

Table 1 List of notations

符号	描述
$S$	给定的文本句子
$R$	预先定义的关系类型集合
$E$	预先定义的实体类型集合
$w / w i$	句子中的单词/第 $i$ 个单词
$r$	关系类型
$e$	实体类型
$h$	头实体
$t$	尾实体
$e$	嵌入向量
$h / h i$	隐藏状态向量/ $i$ 时刻隐藏状态向量
$W / W i$	参数矩阵/第 $i$ 层的参数矩阵
$b / b i$	参数向量/第 $i$ 层的参数向量

表1 常用符号描述

Table 1 List of notations

符号	描述
$S$	给定的文本句子
$R$	预先定义的关系类型集合
$E$	预先定义的实体类型集合
$w / w i$	句子中的单词/第 $i$ 个单词
$r$	关系类型
$e$	实体类型
$h$	头实体
$t$	尾实体
$e$	嵌入向量
$h / h i$	隐藏状态向量/ $i$ 时刻隐藏状态向量
$W / W i$	参数矩阵/第 $i$ 层的参数矩阵
$b / b i$	参数向量/第 $i$ 层的参数向量

图2 卡片金字塔模型

Fig.2 Card pyramid model

图3 文献[37]的概率图模型

Fig.3 Probability graph model of Ref. [37]

图4 文献[39]的表格标注方法

Fig.4 Table annotation method of Ref. [39]

表2 基于特征工程的联合抽取方法总结

Table 2 Summary of joint extraction methods based on feature engineering

类型	优点	缺点	文献	描述
整数线性规划	通用性和灵活性,线性公式可以表示任意约束类型	耗时,子任务间的交互性较低	[30-31]	根据多个局部分类器的结果求解全局最优分配策略
卡片金字塔模型	通过图结构实现全局和局部信息交互,使得抽取结果更准确	由多个局部模型构成,结构复杂	[33]	构造图结构,将联合抽取转换为图节点标注问题
概率图模型	充分利用子任务间的交互,灵活地融入其他类型特征	需要计算大量概率分布	[36-37]	将实体关系表示成概率图,求解实体关系的最大后验概率
结构化预测	建立单一的联合学习模型	直接进行结构预测,计算复杂度高	[38-39]	采用单一的图或表结构表示实体关系,模型直接预测候选结构

图5 基于神经网络的联合抽取

Fig.5 Joint extraction based on neural network

图6 基于依赖树的联合抽取模型

Fig.6 Joint extraction model based on dependency tree

表3 实体对映射到关系模型总结

Table 3 Summary of mapping entity pairs to relationship models

文献	模型架构	描述
[41]	双向LSTM+依赖树	两个子任务都采用双向LSTM网络和前馈神经网络实现
[44]	双向LSTM+卷积神经网络	卷积神经网络抽取关系类型时融入了实体间的局部句子信息
[45]	双向LSTM	两个子任务均采用LSTM网络实现
[46]	双向LSTM+CRF	采用CRF获得存在候选关系的实体,减少了冗余实体的影响
[47]	双向LSTM+自注意力机制	将每个关系类型当作独立的子空间,用自注意力机制获取更细粒度的语义关联
[48]	双向LSTM+双向GCN	第一阶段预测出所有实体对,第二阶段构造带权重的GCN预测关系
[49]	双向LSTM+注意力机制	用注意力机制获取所有可能的跨度,跨度信息通过前馈神经网络实现关系抽取
[50]	双向LSTM+注意力机制	跨度类型信息和关系信息采用集束搜索的方式进行评估
[51]	双向LSTM+动态跨度图	采用动态跨度图传递上下文信息,进一步丰富跨度信息
[52]	BERT+动态跨度图	采用BERT编码器,提升抽取跨度特征的准确性
[53]	BERT+前馈神经网络	在BERT编码器上进行轻量级推理,并采用负采样降低训练复杂度
[54]	BERT+注意力机制	用注意力机制实现跨度的表示,融入了局部和整体的语义信息实现关系抽取
[55]	表格填充+循环神经网络	将实体关系信息的表格转换成信息,用循环神经网络抽取实体关系信息
[56]	表格填充+双向LSTM	表格填充过程中设计标注打分函数,获得全局最优的序列标注
[57]	GRU+LSTM	两个子任务都采用GRU,用LSTM学习参数层动态交互信息
[58]	双向LSTM+依赖树	采用强化学习方法,增强两个子任务交互性
[60]	双向LSTM+卷积神经网络	设计最小化风险的全局损失函数,增强两个子任务交互性
[61]	双向LSTM+GCN	用序列标注识别实体后,将实体类型和关系类型构造成二分图进行联合推理

表3 实体对映射到关系模型总结

Table 3 Summary of mapping entity pairs to relationship models

文献	模型架构	描述
[41]	双向LSTM+依赖树	两个子任务都采用双向LSTM网络和前馈神经网络实现
[44]	双向LSTM+卷积神经网络	卷积神经网络抽取关系类型时融入了实体间的局部句子信息
[45]	双向LSTM	两个子任务均采用LSTM网络实现
[46]	双向LSTM+CRF	采用CRF获得存在候选关系的实体,减少了冗余实体的影响
[47]	双向LSTM+自注意力机制	将每个关系类型当作独立的子空间,用自注意力机制获取更细粒度的语义关联
[48]	双向LSTM+双向GCN	第一阶段预测出所有实体对,第二阶段构造带权重的GCN预测关系
[49]	双向LSTM+注意力机制	用注意力机制获取所有可能的跨度,跨度信息通过前馈神经网络实现关系抽取
[50]	双向LSTM+注意力机制	跨度类型信息和关系信息采用集束搜索的方式进行评估
[51]	双向LSTM+动态跨度图	采用动态跨度图传递上下文信息,进一步丰富跨度信息
[52]	BERT+动态跨度图	采用BERT编码器,提升抽取跨度特征的准确性
[53]	BERT+前馈神经网络	在BERT编码器上进行轻量级推理,并采用负采样降低训练复杂度
[54]	BERT+注意力机制	用注意力机制实现跨度的表示,融入了局部和整体的语义信息实现关系抽取
[55]	表格填充+循环神经网络	将实体关系信息的表格转换成信息,用循环神经网络抽取实体关系信息
[56]	表格填充+双向LSTM	表格填充过程中设计标注打分函数,获得全局最优的序列标注
[57]	GRU+LSTM	两个子任务都采用GRU,用LSTM学习参数层动态交互信息
[58]	双向LSTM+依赖树	采用强化学习方法,增强两个子任务交互性
[60]	双向LSTM+卷积神经网络	设计最小化风险的全局损失函数,增强两个子任务交互性
[61]	双向LSTM+GCN	用序列标注识别实体后,将实体类型和关系类型构造成二分图进行联合推理

图7 分层级的强化学习框架

Fig.7 Hierarchical reinforcement learning framework

表4 共享参数模型总结

Table 4 Summary of shared parameter model

类型	优点	缺点	方法	文献	描述
实体对映射到关系	存在大量成熟有效的实体命名识别和关系抽取模型,通过共享参数容易实现联合抽取	存在的冗余实体对提高了错误率;难以高效解决关系重叠的问题	循环神经网络为主体	[41,45-47,55-58]	命名实体识别和关系抽取都采用循环神经网络设计
			混合模型	[44,48,60-61]	命名实体识别一般采用循环神经网络,关系抽取采用卷积神经网络、GCN等
			基于跨度	[49-54]	直接对实体跨度建模,能够有效解决实体嵌套的问题
头实体映射到关系、尾实体	加强实体类型信息和关系类型信息间的交互;有效解决关系重叠的问题	识别候选头实体和头实体映射到关系、尾实体两个过程的方法尚不成熟,设计相对复杂	循环神经网络为主体	[62-65]	用循环神经网络融入上下文信息,识别出实体后用注意力方法识别关系和另一个实体
			Transformer架构	[66-67]	用Transformer提取更深层次的特征信息,采用指针网络抽取关系三元组
			多轮问答方法	[68-69]	问题中融入实体类型等先验信息,用机器阅读理解的方法抽取关系三元组
关系映射到头实体、尾实体	减少了冗余信息的抽取;从设计上容易解决关系重叠的问题	识别候选关系类型的难度大,设计相对复杂	循环神经网络为主体	[70-72]	使用循环神经网络编码后,首先解码得到关系类型,再根据关系类型抽取对应实体信息
关系映射到头实体、尾实体	减少了冗余信息的抽取;从设计上容易解决关系重叠的问题	识别候选关系类型的难度大,设计相对复杂	设计两个编码器	[73]	两个编码器分别编码实体信息和关系类型信息,通过前馈神经网络进行预测

图8 序列标注方案

Fig.8 Sequence annotation scheme

表5 关系类型分类示例

Table 5 Example of relationship type classification

实体关系	文本	实体关系
Normal	水浒传的作者是施耐庵	作者水浒传施耐庵
EPO	北京,有着灿烂的文化、悠久的历史和丰富的古迹,是中国的首都	首都中国北京包含
SEO	刘备的二弟,关羽,温酒斩华雄,一战成名。

图9 Sequence-to-Sequence模型对比

Fig.9 Sequence-to-Sequence model comparison

表6 基于神经网络的联合抽取模型

Table 6 Joint extraction model based on neural network

类型	优点	缺点	方法	文献	描述
共享参数	不同子任务在构建模型时能抽取丰富的特征信息	不同子任务之间信息交互不够充分	实体对映射到关系	[41,44-61]	先进行命名实体识别,再根据其结果实现关系抽取
			头实体映射到关系、尾实体	[62-69]	先识别头实体,根据头实体抽取相应关系和尾实体
			关系映射到头实体、尾实体	[70-73]	先抽取关系,依据关系类型建模到实体对的映射
联合解码	设计统一的解码器,实体关系信息得以充分交互	设计复杂的解码架构使得局部特征抽取不充分	序列标注	[74-75]	设计复杂的标注方案融入实体、关系信息后解码序列
联合解码	设计统一的解码器,实体关系信息得以充分交互	设计复杂的解码架构使得局部特征抽取不充分	Sequence-to-Sequence	[78,80-84]	解码器根据编码得到的语义向量依次产生关系三元组

表7 ACE2004数据集

Table 7 ACE2004 dataset

实体类型	关系类型
Person	PHYS
Organization	PER-SOC
Geographical Entities	PER/ORG-AFF
Location	ART
Facility	EMP-ORG
Weapon	GPE-AFF
Vehicle	DISC

表8 CoNLL04数据集

Table 8 CoNLL04 dataset

实体类型及数量	关系类型及数量
	406 Located In
1 685 Person	394 Work For
1 968 Location	451 OrgBased In
978 Organization	521 Live In
705 Other	268 Kill
	17 007 None

表9 实体关系抽取数据集总结

Table 9 Summary of entity and relation extraction datasets

数据集	实体种类	关系种类	规模	数据来源	下载网址
ACE2004	7	7	6 800	语言数据联盟	https://catalog.ldc.upenn.edu/LDC2005T09
ACE2005	7	6	10 500	语言数据联盟	https://catalog.ldc.upenn.edu/LDC2006T06
SemEval-2010 Task 8	—	9	10 700	WordNet等	https://semeval2.fbk.eu/semeval2.php?location=data
CoNLL04	4	6	1 400	国际文本信息检索会议	https://cogcomp.seas.upenn.edu/page/resource_view/43
ADE	2	1	6 800	美国国家医学图书馆	https://github.com/lavis-nlp/spert/tree/master/scripts
NTY	3	24	66 200	纽约时报	https://github.com/xiangrongzeng/copy_re
WebNLG	—	246	6 200	DBpedia	https://github.com/weizhepei/CasRel/tree/master/data/WebNLG

表10 混淆矩阵

Table 10 Confusion matrix

真实情况	预测结果
真实情况	正类	反类
正类	TP	FN
反类	FP	TN

图10 NYT数据集上的评测结果

Fig.10 Evaluation results on NYT dataset

图11 WebNLG数据集上的评测结果

Fig.11 Evaluation results on WebNLG dataset

表11 有监督数据集上评测结果

Table 11 Evaluation results on supervised datasets %

数据集	模型	命名实体识别 $F 1$	关系抽取 $F 1$
ACE2004	Li^[38]	79.7	45.3
	Katiyar^[62]	79.6	45.7
	Bekoulis^[63]	81.2	47.1
	Bekoulis^[64]	81.6	47.5
	SPTree^[41]	81.8	48.4
	Li^[68]	83.6	49.4
	DyGIE^[51]	87.4	59.7^*
	Wang^[73]	88.6	59.6
ACE2005	Li^[38]	80.8	49.5
	SPTree^[41]	83.4	55.6
	Katiyar^[62]	82.6	53.6
	Zhang^[56]	83.5	57.5
	Sun^[60]	83.6	59.6
	Sun^[61]	84.2	59.1
	Li^[68]	84.8	60.2
	Zhao^[69]	85.7	62.3
	Dixit^[49]	86.0	62.8^*
	DyGIE^[51]	88.4	63.2^*
	Wadden^[52]	88.6	63.4^*
	Wang^[73]	89.5	64.3
	SPAN^[54]	89.6	65.2
CoNLL04	Miwa^[39]	80.7	61.0
	Bekoulis^[63]	83.6	62.0
	Bekoulis^[64]	83.9	62.0
	Zhang^[56]	85.6	67.8
	Li^[68]	87.8	68.9
	SpERT^[53]	88.9	71.5
	Zhao^[69]	88.9	71.9
	Wang^[73]	90.1	73.6
	SPAN^[54]	90.2	74.3
ADE	Li^[45]	84.6	71.4
	Bekoulis^[63]	86.4	74.6
	Bekoulis^[64]	86.7	75.5
	SpERT^[53]	89.3	79.2
	Wang^[73]	89.7	80.1
	SPAN^[54]	90.6	80.7

表11 有监督数据集上评测结果

Table 11 Evaluation results on supervised datasets %

数据集	模型	命名实体识别 $F 1$	关系抽取 $F 1$
ACE2004	Li^[38]	79.7	45.3
	Katiyar^[62]	79.6	45.7
	Bekoulis^[63]	81.2	47.1
	Bekoulis^[64]	81.6	47.5
	SPTree^[41]	81.8	48.4
	Li^[68]	83.6	49.4
	DyGIE^[51]	87.4	59.7^*
	Wang^[73]	88.6	59.6
ACE2005	Li^[38]	80.8	49.5
	SPTree^[41]	83.4	55.6
	Katiyar^[62]	82.6	53.6
	Zhang^[56]	83.5	57.5
	Sun^[60]	83.6	59.6
	Sun^[61]	84.2	59.1
	Li^[68]	84.8	60.2
	Zhao^[69]	85.7	62.3
	Dixit^[49]	86.0	62.8^*
	DyGIE^[51]	88.4	63.2^*
	Wadden^[52]	88.6	63.4^*
	Wang^[73]	89.5	64.3
	SPAN^[54]	89.6	65.2
CoNLL04	Miwa^[39]	80.7	61.0
	Bekoulis^[63]	83.6	62.0
	Bekoulis^[64]	83.9	62.0
	Zhang^[56]	85.6	67.8
	Li^[68]	87.8	68.9
	SpERT^[53]	88.9	71.5
	Zhao^[69]	88.9	71.9
	Wang^[73]	90.1	73.6
	SPAN^[54]	90.2	74.3
ADE	Li^[45]	84.6	71.4
	Bekoulis^[63]	86.4	74.6
	Bekoulis^[64]	86.7	75.5
	SpERT^[53]	89.3	79.2
	Wang^[73]	89.7	80.1
	SPAN^[54]	90.6	80.7

参考文献 86

[1]	GOLSHAN P N, DASHTI H R, AZIZI S, et al. A study of recent contributions on information extraction[J]. arXiv: 1803. 05667, 2018.
[2]	FREITAG D. Machine learning for information extraction in informal domains[J]. Machine Learning, 2000, 39(2):169-202. DOI URL
[3]	刘春梅, 郭岩, 俞晓明, 等. 针对开源论坛网页的信息抽取研究[J]. 计算机科学与探索, 2017, 11(1):114-123.
	LIU C M, GUO Y, YU X M, et al. Information extraction research aimed at open sourceweb pages[J]. Journal of Frontiers of Computer Science and Technology, 2017, 11(1):114-123.
[4]	薛丽娟, 席梦隆, 王梦婕, 等. 基于规则推理引擎的实体关系抽取研究[J]. 计算机科学与探索, 2016, 10(9):1310-1319.
	XUAN L J, XI M L, WANG M J, et al. Entity relation ex-traction based on rule inference engine[J]. Journal of Frontiers of Computer Science and Technology, 2016, 10(9):1310-1319.
[5]	甘丽新, 万常选, 刘德喜, 等. 基于句法语义特征的中文实体关系抽取[J]. 计算机研究与发展, 2016, 53(2):284-302.
	GAN L X, WAN C Y, LIU D C, et al. Chinese named entity relation extraction based on syntactic and semantic features[J]. Journal of Computer Research and Development, 2016, 53(2):284-302.
[6]	BENDER O, OCH F J, NEY H. Maximum entropy models for named entity recognition[C]// Proceedings of the 17th Conference on Natural Language Learning at HLT-NAACL, Edmonton, May 31-Jun 1, 2003. Stroudsburg: ACL, 2003: 148-151.
[7]	BIKEL D M, SCHWARTZ R, WEISCHEDEL R M. An algo-rithm that learns what’s in a name[J]. Machine Learning, 1999, 34(1):211-231. DOI URL
[8]	WANG B, LU W, WANG W, et al. A neural transition-based model for nested mention recognition[C]// Proceedings of the 2018 Conference on Empirical Methods in Natural Lan-guage Processing, Brussels, Oct 31-Nov 4, 2018. Strouds-burg: ACL, 2018: 1011-1017.
[9]	ZELENKO D, AONE C, RICHARDELLA A. Kernel methods for relation extraction[J]. Journal of Machine Learning Research, 2003(3):1083-1106.
[10]	ZHOU G D, SU J, ZHANG J, et al. Exploring various knowledge in relation extraction[C]// Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics, Michigan, Jun 25-30, 2005. Stroudsburg: ACL, 2005: 427-434.
[11]	CHAN Y S, ROTH D. Exploiting syntactico-semantic struc-tures for relation extraction[C]// Proceedings of the 49th Ann-ual Meeting of the Association for Computational Lingui-stics, Portland, Jun 19-24, 2011. Stroudsburg: ACL, 2011: 551-560.
[12]	MILLKER S, FOX H, RAMSHAW L, et al. A novel use of statistical parsing to extract information from text[C]// Pro-ceedings of the 6th Applied Natural Language Processing Conference, Seattle, Apr 29-May 4, 2000. Stroudsburg: ACL, 2000: 226-233.
[13]	陈宇, 郑德权, 赵铁军. 基于Deep Belief Nets的中文名实体关系抽取[J]. 软件学报, 2012, 23(10):2572-2585.
	CHEN Y, ZHENG D H, ZHAO T J. Chinese relation extrac-tion based on Deep Belief Nets[J]. Journal of Software, 2012, 23(10):2572-2585.
[14]	鄂海红, 张文静, 肖思琪, 等. 深度学习实体关系抽取研究综述[J]. 软件学报, 2019, 30(6):1793-1818.
	E H H, ZHANG W J, XIAO S Q, et al. oint entity rela-tionship extraction based on deep learning[J]. Journal of Software, 2019, 30(6):1793-1818.
[15]	李冬梅, 张扬, 李东远, 等. 实体关系抽取方法研究综述[J]. 计算机研究与发展, 2020, 57(7):1424-1448.
	LI D M, ZHANG Y, LI D Y, et al. A survey of entity relation extraction methods[J]. Journal of Computer Research and Development, 2020, 57(7):1424-1448.
[16]	PAWAR S, PALSHIKAR G K, BHATTACHARYYA P. Etrac-tion: a survey[J]. arXiv: 1712. 05191, 2017.
[17]	KONSTANTIONVA N. Review of relation extraction me-thods: what is new out there?[C]// Proceedings of the 3rd International Conference on Analysis of Images, Social Net-works and Texts, Yekaterinburg, Apr 10-12, 2014. Cham: Springer, 2014: 15-28.
[18]	KUMAR S. A survey of deep learning methods for relation extraction[J]. arXiv: 1705. 03645, 2017.
[19]	ZHANG Q Q, CHEN M D, LIU L Z. A review on entity re-lation extraction[C]// Proceedings of the 2017 2nd Interna-tional Conference on Mechanical, Control and Computer Engineering, Harbin, Dec 8-10, 2017. Piscataway: IEEE, 2017: 178-183.
[20]	PAWAR S, BHATTACHARYYA P, PALSHIKAR G K. Tech-niques for jointly extracting entities and relations: a survey[J]. arXiv: 2103. 06118, 2021.
[21]	ELMAN J L. Finding structure in time[J]. Cognitive Science, 1990, 14(2):179-211. DOI URL
[22]	HOCHREITER S, SCHMIDHUBER J. Long short-term me-mory[J]. Neural Computation, 1997, 9(8):1735-1780. DOI URL
[23]	CHUNG J Y, GULCEHRE C, CHO K, et al. Empirical eva-luation of gated recurrent neural networks on sequence mode-ling[J]. arXiv: 1412. 3555, 2014.
[24]	KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[J]. arXiv: 1609. 02907, 2006.
[25]	DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understan-ding[C]// Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Lin-guistics: Human Language Technologies, Minneapolis, Jun 2-7, 2019. Stroudsburg: ACL, 2019: 4171-4186.
[26]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]// Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Pro-cessing Systems 2017, Long Beach, Dec 4-9, 2017: 5998-6008.
[27]	FLORIAN R, ITTYCHERIAH A, JING H, et al. Named entity recognition through classifier combination[C]// Pro-ceedings of the 17th Conference on Natural Language Learning at HLT-NAACL, Edmonton, May 31-Jun 1, 2003. Stroudsburg: ACL, 2003: 168-171.
[28]	MIWA M, RUNE S, YUSUKE M, et al. A rich feature vector for protein-protein interaction extraction from multiple cor-pora[C]// Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing, Singapore, Aug 6-7, 2009. Stroudsburg: ACL, 2009: 121-130.
[29]	DANTZIG G B. Reminiscences about the origins of linear pro-gramming[J]. Operations Research Letters, 1982, 1(2):43-48. DOI URL
[30]	ROTH D, YIN W T. A linear programming formulation for global inference in natural language tasks[C]// Proceedings of the 8th Conference on Computational Natural Language Lear-ning, Boston, May 6-7, 2004. Stroudsburg: ACL, 2004: 1-8.
[31]	YANG B, CARDIE C. Joint inference for fine-grained opinion extraction[C]// Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, Sofia, Aug 4-9, 2013. Stroudsburg: ACL, 2013: 1640-1649.
[32]	LAFFERT J, MCCALLUM A, PEREIRA F C. Conditional random fields: probabilistic models for segmenting and labeling sequence data[C]// Proceedings of the 18th Interna-tional Conference on Machine Learning, Williamstown, Jun 28-Jul 1, 2001. San Francisco: Morgan Kaufmann, 2001: 282-289.
[33]	KATE R, NOONEY R. A joint entity and relation extraction using card-pyramid parsing[C]// Proceedings of the 14th Con-ference on Computational Natural Language Learning, Uppsala, Jul 15-16, 2010. Stroudsburg: ACL, 2010: 203-212.
[34]	CORTES C, VAPNIK V. Support-vector networks[J]. Ma-chine Learning, 1995, 20(3):273-297.
[35]	GHAHRAMANI Z. Probabilistic machine learning and ar-tificial intelligence[J]. Nature, 2015, 521(7553):452-459. DOI URL
[36]	YU X F, LAM W. Jointly identifying entities and extracting relations in encyclopedia text via a graphical model approach[C]// Proceedings of the 23rd International Conference on Computational Linguistics, Beijing, Aug 23-27, 2010. Strouds-burg: ACL, 2010: 1399-1407.
[37]	SINGH S, RIEDEL S, MARTIN B, et al. Joint inference of entities, relations, and coreference[C]// Proceedings of the 2013 Workshop on Automated Knowledge Base Construction, San Francisco, Oct 27-28, 2013. New York: ACM, 2013: 1-6.
[38]	LI Q, JI H. Incremental joint extraction of entity mentions and relations[C]// Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, Balti-more, Jun 22-27, 2014. Stroudsburg: ACL, 2014: 402-412.
[39]	MIWA M, SASAKI Y. Modeling joint entity and relation extraction with table representation[C]// Proceedings of the 2014 Conference on Empirical Methods in Natural Lan-guage Processing, Doha, Oct 25-29, 2014. Stroudsburg: ACL, 2014: 1858-1869.
[40]	HINTON G E, SALAKHUTDINOV S S. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786):504-407. DOI URL
[41]	MIWA M, BANSAL M. End-to-end relation extraction using LSTMs on sequences and tree structures[C]// Proceedings of the 54th Annual Meeting of the Association for Com-putational Linguistics, Berlin, Aug 7-12, 2016. Stroudsburg: ACL, 2016: 1105-1116.
[42]	XU K, FENG Y, HUANG S, et al. Semantic relation classifi-cation via convolutional neural networks with simple nega-tive sampling[C]// Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, Lis-bon, Sep 17-21, 2015. Stroudsburg: ACL, 2015: 536-540.
[43]	WERBOS P J. Backpropagation through time: what it does and how to do it[J]. Proceedings of the IEEE, 1990, 78(10):1550-1560. DOI URL
[44]	ZHENG S C, HAO Y X, LU D Y, et al. Joint entity and relation extraction based on a hybrid neural network[J]. Neurocomputing, 2017, 267:59-66.
[45]	LI F, ZHANGE M, FU G, et al. A neural joint model for entity and relation extraction from biomedical text[J]. BMC Bioinformatics, 2017, 18(1):1-11.
[46]	TAN Z, ZHAO X, WANG W, et al. Jointly extracting multi-ple triplets with multilayer translation constraints[C]// Pro-ceedings of the 33rd AAAI Conference on Artificial Intelli-gence, Honolulu, Jan 27-Feb 1, 2019. Menlo Park: AAAI, 2019: 7080-7087.
[47]	LIU J, CHEN S W, WANG B Q, et al. Attention as relation: learning supervised multi-head self-attention for relation ex-traction[C]// Proceedings of the 29th International Joint Conference on Artificial Intelligence, Yokohama, Jan 7-15, 2021. San Francisco: Morgan Kaufmann, 2021: 3787-3793.
[48]	FU T J, LI P H, MA W Y. GraphRel: modeling text as rela-tional graphs for joint entity and relation extraction[C]// Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, Florence, Jul 28-Aug 2, 2019. Stroudsburg: ACL, 2019: 1409-1418.
[49]	DIXIT K, AL-ONAIZAN Y. Span-level model for relation extraction[C]// Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, Florence, Jul 28-Aug 2, 2019. Stroudsburg: ACL, 2019: 5308-5314.
[50]	LUAN Y, HE L Y, OSTENDORF M, et al. Multi-task iden-tification of entities, relations,coreference for scien-tific knowledge graph construction[J]. arXiv: 1808. 09602, 2018.
[51]	LUAN Y, WADDEN D, HE L H, et al. A general frame-work for information extraction using dynamic span graphs[C]// Proceedings of the 2019 Conference of the North Ame-rican Chapter of the Association for Computational Lin-guistics: Human Language Technologies, Minneapolis, Jun 2-7, 2019. Stroudsburg: ACL, 2019: 3036-3046.
[52]	WADDEN D, WENNBERG U, LUAN L, et al. Entity, rela-tion, and event extraction with contextualized span repre-sentations[C]// Proceedings of the 2019 Conference on Em-pirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, Hong Kong, China, Nov 3-7, 2019. Stroudsburg: ACL, 2019: 5784-5789.
[53]	EBERTS M, ULGES A. Span-based joint entity and relation extraction with transformer pre-training[J]. arXiv: 1909. 07755, 2019.
[54]	JI B, YU J, LI S S, et al. Span-based joint entity and relation extraction with attention-based span-specific and contextual semantic representations[C]// Proceedings of the 28th Inter-national Conference on Computational Linguistics, Barce-lona, Dec 8-13, 2020. Praha: ICCL, 2020: 88-99.
[55]	GUPTA P, SCHUTZE H, ANDRASSY B. Table filling multi-task recurrent neural network for joint entity and relation extraction[C]// Proceedings of the 26th International Confe-rence on Computational Linguistics, Osaka, Dec 11-16, 2016. Stroudsburg: ACL, 2016: 2537-2547.
[56]	ZHANG M, ZHANG Y, FU G. End-to-end neural relation extraction with global optimization[C]// Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, Copenhagen, Sep 9-11, 2017. Stroudsburg: ACL, 2017: 1730-1740.
[57]	SUN K, ZHANG R, MENSAH S, et al. Recurrent interac-tion network for jointly extracting entities and classifying rela-tions[C]// Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, Nov 16-20, 2020. Stroudsburg: ACL, 2020: 3722-3732.
[58]	FENG Y, ZHANG H J, HAO W N, et al. Joint extraction of entities and relations using reinforcement learning and deep learning[J]. Computational Intelligence and Neuroscience, 2017, 7643065:1-11.
[59]	KAELBING L P, LTTMAN M L, MOORE A W. Reinforce-ment learning: a survey[J]. Journal of Artificial Intelligence Research, 1996, 4:237-285. DOI URL
[60]	SUN C Z, WU Y, LAN M, et al. Extracting entities and relations with joint minimum risk training[C]// Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Oct 31-Nov 4, 2018. Strouds-burg: ACL, 2018: 2256-2265.
[61]	SUN C Z, GONG Y Y, WU Y B, et al. Joint type inference on entities and relations via graph convolutional networks[C]// Proceedings of the 57th Annual Meeting of the Asso-ciation for Computational Linguistics, Florence, Jul 28-Aug 2, 2019. Stroudsburg: ACL, 2019: 1361-1370.
[62]	KATIYAR A, CARDIE C. Going out on a limb: joint extrac-tion of entity mentions and relations without dependency trees[C]// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, Vancouver, Jul 30-Aug 4, 2017. Stroudsburg: ACL, 2017: 917-928.
[63]	BEKOULIS G, DELEU J, DEMEESTER T, et al. Joint entity recognition and relation extraction as a multi-head selection pro-blem[J]. Expert Systems with Applications, 2018, 114:34-45. DOI URL
[64]	BEKOULIS G, DELEU J, DEMEESTER T, et al. Adversarial training for multi-context joint entity and relation extraction[C]// Proceedings of the 2018 Conference on Empirical Me-thods in Natural Language Processing, Brussels, Oct 31-Nov 4, 2018. Stroudsburg: ACL, 2018: 2830-2836.
[65]	YU B, ZHANG Z Y, SHU X B, et al. Joint extraction of entities and relations based on a novel decomposition stra-tegy[J]. arXiv: 1909. 04273, 2019.
[66]	WEI Z P, SU J L, WANG Y, et al. A novel cascade binary tagging framework for relational triple extraction[C]// Pro-ceedings of the 58th Annual Meeting of the Association for Computational Linguistics, Jul 5-10, 2020. Stroudsburg: ACL, 2020: 1476-1488.
[67]	WANG Y C, YU B W, ZHANG Y Y, et al. TPLinker: single-stage joint extraction of entities and relations through token pair linking[C]// Proceedings of the 28th International Confe-rence on Computational Linguistics, Barcelona, Dec 8-13, 2020. Praha: ICCL, 2020: 1572-1582.
[68]	LI X, YIN F, SUN Z, et al. Entity-relation extraction as multi-turn question answering[C]// Proceedings of the 57th Annual Meeting of the Association for Computational Lin-guistics, Florence, Jul 28-Aug 2, 2019. Stroudsburg: ACL, 2019: 1340-1350.
[69]	ZHAO T, YAN Z, CAO Y, et al. Asking effective and diverse questions: a machine reading comprehension based frame-work for joint entity-relation extraction[C]// Proceedings of the 29th International Joint Conference on Artificial In-telligence, Yokohama, Jan 7-15, 2021. San Francisco: Mor-gan Kaufmann, 2021: 3948-3954.
[70]	TAKANOBU R, ZHANG T, LIU J, et al. A hierarchical frame-work for relation extraction with reinforcement learning[C]// Proceedings of the 33rd AAAI Conference on Arti-ficial Intelligence, Honolulu, Jan 27-Feb 1, 2019. Menlo Park: AAAI, 2019: 7072-7079.
[71]	ZHOU P, ZHENG S, XU J, et al. Joint extraction of multi-ple relations and entities by using a hybrid neural network[C]// Proceedings of the 16th China National Conference on Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data, Nan-jing, Oct 13-15, 2017. Cham: Springer, 2017: 135-146.
[72]	YUAN Y, ZHOU X, PAN S, et al. A relation-specific at-tention network for joint entity and relation extraction[C]// Proceedings of the 29th International Joint Conferences on Artificial Intelligence, Yokohama, Jan 7-15, 2021. San Fran-cisco: Morgan Kaufmann, 2020: 4054-4060.
[73]	WANG J, WEI L. Two are better than one: joint entity and relation extraction with table-sequence encoders[C]// Pro-ceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, Nov 16-20, 2020. Stroudsburg: ACL, 2020: 1706-1721.
[74]	ZHENG S, WANG F, BAO H, et al. Joint extraction of enti-ties and relations based on a novel tagging scheme[C]// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, Vancouver, Jul 30-Aug 4, 2017. Stroudsburg: ACL, 2017: 1227-1236.
[75]	DAI D, XIAO X, LYU Y, et al. Joint extraction of entities and overlapping relations using position-attentive sequence labeling[C]// Proceedings of the 33rd AAAI Conference on Artificial Intelligence, Honolulu, Jan 27-Feb 1, 2019. Menlo Park: AAAI, 2019: 6300-6308.
[76]	CHO K, MERRIENBOER B V, GULCEHRE C, et al. Lear-ning phrase representations using RNN encoder-decoder for statistical machine translation[C]// Proceedings of the 2014 Conference on Empirical Methods in Natural Language Pro-cessing, Doha, Oct 25-29, 2014. Stroudsburg: ACL, 2014: 1724-1734.
[77]	SUTAKEVER I, VINYALS O, LE Q V. Sequence to se-quence learning with neural networks[C]// Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014, Montreal, Dec 8-13, 2014: 3104-3112.
[78]	ZENG X R, ZENG D, HE S, et al. Extracting relational facts by an end-to-end neural model with copy mechanism[C]// Proceedings of the 56th Annual Meeting of the Asso-ciation for Computational Linguistics, Melbourne, Jul 15-20. Stroudsburg: ACL, 2018: 506-514.
[79]	SCHUSTER M, PALIWAL K K. Bidirectional recurrent neu-ral networks[J]. IEEE Transactions on Signal Processing, 1997, 45(11):2673-2681. DOI URL
[80]	ZENG X R, HE S Z, ZENG D J, et al. Learning the ex-traction order of multiple relational facts in a sentence with reinforcement learning[C]// Proceedings of the 2019 Con-ference on Empirical Methods in Natural Language Proces-sing and the 9th International Joint Conference on Natural Language Processing, Hong Kong, China, Nov 3-7, 2019. Stroudsburg: ACL, 2019: 367-377.
[81]	ZENG D, ZHANG H, LIU Q. CopyMTL: copy mechanism for joint extraction of entities and relations with multi-task learning[C]// Proceedings of the 34th AAAI Conference on Artificial Intelligence, New York, Feb 7-12, 2020. Menlo Park: AAAI, 2020: 9507-9514.
[82]	PANG Y, LIU J, LIU L, et al. A deep neural network model for joint entity and relation extraction[J]. IEEE Access, 2019, 7:179143-179150. DOI URL
[83]	NAYAK T, NG H. Effective modeling of encoder-decoder architecture for joint entity and relation extraction[C]// Pro-ceedings of the 34th AAAI Conference on Artificial In-telligence, New York, Feb 7-12, 2020. Menlo Park: AAAI, 2020: 8528-8535.
[84]	SUI D, CHEN Y, LIU K, et al. Joint entity and relation extraction with set prediction networks[J]. arXiv: 2011. 01675, 2020.
[85]	GATT A, KRAHMER E. Survey of the state of the art in natural language generation: core tasks, applications and evaluation[J]. Journal of Artificial Intelligence Research, 2018, 61:65-170. DOI URL
[86]	REN X, WU Z, HE W, et al. CoType: joint extraction of typed entities and relations with knowledge bases[C]// Pro-ceedings of the 26th International Conference on World Wide Web, Perth, Apr 3-7, 2017. New York: ACM, 2017: 1015-1024.

有监督实体关系联合抽取方法研究综述

Survey of Supervised Joint Entity Relation Extraction Methods

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 22

参考文献 86

相关文章 15

编辑推荐

Metrics

[1]	安凤平, 李晓薇, 曹翔. 权重初始化-滑动窗口CNN的医学图像分类[J]. 计算机科学与探索, 2022, 16(8): 1885-1897.
[2]	黄浩, 葛洪伟. 强化类间区分的深度残差表情识别网络[J]. 计算机科学与探索, 2022, 16(8): 1842-1849.
[3]	于慧琳, 陈炜, 王琪, 高建伟, 万怀宇. 使用子图推理实现知识图谱关系预测[J]. 计算机科学与探索, 2022, 16(8): 1800-1808.
[4]	李玉轩, 洪学海, 汪洋, 唐正正, 班艳. 引入激活加权策略的分组排序学习方法[J]. 计算机科学与探索, 2022, 16(7): 1594-1602.
[5]	张雁操, 赵宇海, 史岚. 融合图注意力的多特征链接预测算法[J]. 计算机科学与探索, 2022, 16(5): 1096-1106.
[6]	欧阳柳, 贺禧, 瞿绍军. 全卷积注意力机制神经网络的图像语义分割[J]. 计算机科学与探索, 2022, 16(5): 1136-1145.
[7]	程卫月, 张雪琴, 林克正, 李骜. 融合全局与局部特征的深度卷积神经网络算法[J]. 计算机科学与探索, 2022, 16(5): 1146-1154.
[8]	童敢, 黄立波. Winograd快速卷积相关研究综述[J]. 计算机科学与探索, 2022, 16(5): 959-971.
[9]	裴利沈, 赵雪专. 群体行为识别深度学习方法研究综述[J]. 计算机科学与探索, 2022, 16(4): 775-790.
[10]	卓天天, 桑庆兵. 注意力机制与复合卷积在手写识别中的应用[J]. 计算机科学与探索, 2022, 16(4): 888-897.
[11]	陆仲达, 张春达, 张佳奇, 王子菲, 许军华. 双分支网络的苹果叶部病害识别[J]. 计算机科学与探索, 2022, 16(4): 917-926.
[12]	马金林, 张裕, 马自萍, 毛凯绩. 轻量化神经网络卷积设计研究进展[J]. 计算机科学与探索, 2022, 16(3): 512-528.
[13]	裴利沈, 刘少博, 赵雪专. 人体行为识别研究综述[J]. 计算机科学与探索, 2022, 16(2): 305-322.
[14]	赵山, 罗睿, 蔡志平. 中文命名实体识别综述[J]. 计算机科学与探索, 2022, 16(2): 296-304.
[15]	肖泽管, 陈清亮. 融合多种类型语法信息的属性级情感分析模型[J]. 计算机科学与探索, 2022, 16(2): 395-402.