While traditional systems for Open Information Extraction were statistical and rule-based, recently neural models have been introduced for the task. Our work builds upon CopyAttention, a sequence generation OpenIE model (Cui et. al., 2018). Our analysis reveals that CopyAttention produces a constant number of extractions per sentence, and its extracted tuples often express redundant information.
We present IMoJIE, an extension to CopyAttention, which produces the next extraction conditioned on all previously extracted tuples. This approach overcomes both shortcomings of CopyAttention, resulting in a variable number of diverse extractions per sentence. We train IMoJIE on training data bootstrapped from extractions of several non-neural systems, which have been automatically filtered to reduce redundancy and noise. IMoJIE outperforms CopyAttention by about 18 F1 pts, and a BERT-based strong baseline by 2 F1 pts, establishing a new state of the art for the task.
Boosting Event Extraction with Denoised Structure-to-Text Augmentation
Beijing Institute of Technology(北京理工大学);Ministry of Industry and Information Technology(工业和信息化部);Beijing University of Technology(北京工业大学);Baidu Inc.(百度公司)
A Walk-based Model on Entity Graphs for Relation Extraction
National Centre for Text Mining(国家文本挖掘中心);School of Computer Science, The University of Manchester, United Kingdom(曼彻斯特大学计算机科学学院);Artificial Intelligence Research Center (AIRC), National Institute of Advanced Industrial Science and Technology (AIST), Japan(日本先进工业科学技术研究所人工智能研究中心);Toyota Technological Institute, Nagoya, 468-8511, Japan(丰田工业大学)
Reinforcement Learning-based Dialogue Guided Event Extraction to Exploit Argument Relations
Beijing Advanced Innovation Center for Big Data and Brain Computing, Beihang University(北京航空航天大学大数据与脑机智能高精尖创新中心);Macquarie University(麦考瑞大学);National Computer Network Emergency Response Technical Team/Coordination Center of China(国家计算机网络应急技术处理协调中心);University of Illinois at Chicago(伊利诺伊大学芝加哥分校);University of Leeds(利兹大学)
Document-Level Event Extraction via Human-Like Reading Process
Institute of Information Engineering, Chinese Academy of Sciences(中国科学院信息工程研究所);School of Cyber Security, University of Chinese Academy of Sciences(中国科学院大学网络空间安全学院);Beijing University of Posts and Telecommunications(北京邮电大学)
文档级事件提取(DEE)特别棘手,因为它带来了两个挑战:分散参数和多事件。第一个挑战意味着一个事件记录的参数可以驻留在文档中的不同句子中,而第二个挑战反映了一个文档可以同时包含多个这样的事件记录。本文从人类阅读认知的角度出发,提出了一种基于文本事件的人类阅读启发提取器(Human Reading inspired Extractor for Document Events,HRE),该方法将文本事件分解为粗读和精读两个迭代阶段。具体而言,第一阶段浏览文档以检测事件的发生,第二阶段用于提取特定的事件参数。对于每一个具体的事件角色,精细阅读从句子到字符分层工作,以确定跨句子的论点,从而解决分散的论点问题。同时,通过多轮的粗读探索,发现未检测到的事件,从而解决了多事件问题。实验结果表明,HRE优于以往的竞争方法。
关系抽取是从非结构化文本中抽取知识的重要方法,但需要大量的标注语料。为了减少昂贵的注释工作,半监督学习旨在利用标记和未标记的数据。在本文中,我们回顾和比较了三种典型的方法在半监督RE与深度学习或元学习:自集成,这迫使一致的扰动下,但可能会面临监督不足;自训练,迭代生成伪标签,并重新训练自己与扩大的标记集;双学习,利用一个原始任务和一个双任务,以提供相互反馈。平均教师(Tarvainen和Valpola,2017),LST(Li等人,2019)和DualRE(Lin et al.,2019)为代表,分别阐述了这三种方法的不足。
Joint Bootstrapping Machines for High Confidence Relation Extraction
Siemens AG(西门子股份公司);University of Munich (LMU)(慕尼黑大学(LMU))
用于从文本中提取关系的半监督自举技术迭代地扩展一组初始种子实例。由于缺乏标记数据,自举中的一个关键挑战是语义漂移:如果在迭代期间添加了假阳性实例,则所有后续迭代都将受到污染。我们介绍BREX,一个新的自举方法,防止这种污染的高度有效的信心评估。这是通过联合使用实体和模板种子(而不是在以前的工作中只有一个),通过并行扩展实体和模板,并在每次迭代中相互约束的方式,并通过引入更高质量的模板相似性措施。实验结果表明,BREX在四个关系中实现了比现有技术好0.13(0.87 vs. 0.74)的F1。
Iterative Prompt Refinement for Radiation Oncology Symptom Extraction Using Teacher-Student Large Language Models
Michigan State University(密歇根州立大学);Henry Ford Cancer Institute(亨利·福特癌症研究所);Alexandria University(亚历山大大学);Oakland University(奥克兰大学);Cedars Sinai Medical Center(西达赛奈医疗中心)
Information Extraction from Electricity Invoices with General-Purpose Large Language Models
Centro de Tecnologías de la Imagen (CTIM)(图像技术中心);Instituto Universitario de Cibernética, Empresas y Sociedad (IUCES)(大学网络研究所、企业与社会);University of Las Palmas de Gran Canaria(大加那利岛大学)