刘畅,潘理虎

• 1:天津工业大学计算机科学与技术学院
• 2:太原科技大学计算机科学与技术学院

摘要(Abstract)：

在自然语言处理中，实体关系抽取是一项基本任务，被广泛应用于知识图谱和智能搜索等任务中。中文关系抽取工作中，现有方法整体性能不高，不能有效表示中文包含的信息。提出一个多特征融合的词级注意关系抽取方法MFAN(multi-feature word attention network),并应用于中文关系抽取任务。首先构建一个融合字形、拼音和句法依存辅助特征的卷积神经网络来挖掘中文文本，充分提取其蕴藏的语义信息，再添加词级注意力模块抓取关键词信息，重新分配多特征向量权重。最后基于中文关系抽取数据集，与多个实验进行了对比评估，证明了该方法的有效性。

关键词(KeyWords)： 关系抽取;注意力机制;多特征;卷积网络

Abstract:

Keywords:

基金项目(Foundation): 山西省自然科学基金(201901D111258)

作者(Author): 刘畅,潘理虎

参考文献(References)：

• [1] KAMBHATLA N.Combining lexical,syntactic,and semantic features with maximum entropy models for extracting relations[C]//Processding of ACL.Barcelona,Spain:Association for Computational Linguistics,2004:22-26.
• [2] ZELENKO D,AONE C,RICHARDELLA A.Kernel methods for relation extraction[J].Journal of Machine Learning Research,2003,3(3):1083-1106.
• [3] ZHU H,LIN Y,LIU Z,et al.Graph Neural Networks with Generated Parameters for Relation Extraction [C]//Processding of ACL.Florence,Italy:Association for Computational Linguistics,2019:1331-1339.
• [4] LIU C Y,SUN W B,CHAO W H,et al.Convolution neural network for relation extraction[C]//Processding of ADMA.Berlin,Heidelberg:Springer,2013:231-242.
• [5] MIWA M,BANSAL M.End-to-end relation extraction using LSTMs on Sequences and Tree Structures[J/OL].(2016-01-05).https://arxiv.org/pdf/1601.00770.pdf.
• [6] TSU-JUI FU,PENG-HSUAN LI,WEI-YUN MA.GraphRel:modeling text as relational graphs for joint entity and relation extraction [C]//Processding of ACL.Florence,Italy:Association for Computational Linguistics,2019:1409-1418.
• [7] MINTZ M,BILLS S,SNOW R,et al.Distant supervision for relation extraction without labeled data[C]//Proceedings of ACL.Suntec,Singapore:Association for Computational Linguistics,2009:1003-1011.
• [8] LIU F C,YAN L Y,LIN H Y,et al.Element Intervention for Open Relation Extraction [C]//Processding of ACL.Online:Association for Computational Linguistics,2021:4683-4693.
• [9] YIJUN WANG,CHANGZHI SUN,YUANBIN WU,et al.UNIRE:A Unified Label Space for Entity Relation Extraction[C]//Processding of ACL.Online:Association for Computational Linguistics,2021:220-231.
• [10] KEVIN HUANG,QI P,WANG G T,et al.Entity and Evidence Guided Document-Level Relation Extraction[C]//Processding of RepL4NLP.Online:Association for Computational Linguistics,2021:307-315.
• [11] HU Z K,CAO Y X,HUANG L F,et al.How Knowledge Graph and Attention Help?A Quantitative Analysis into Bag-level Relation Extraction[C]//Processding of ACL.Online:Association for Computational Linguistics,2021:4662-4671.
• [12] 邢百西，赵继舜，刘鹏远.中文关系抽取的句级语言学特征探究[C]//Processding of CCL.Huhhot:Chinese Information Processing Society of China,2021:643-654.
• [13] LI Z R,DING N,LIU Z Y,et al.Chinese Relation Extraction with Multi-Grained Information and External Linguistic Knowledge[C]//Processding of ACL.Florence,Italy:Association for Computational Linguistics,2019:4377-4386.
• [14] QI S S,ZHENG L M,SHANG F Y.Dependency Parsing-based Entity Relation Extraction over Chinese Complex Text[J].ACM Transactions on Asian and Low-Resource Language Information Processing,2021,20(4):1-34.
• [15] MENG Y X,WU W,WANG F,et al.Glyce:Glyph-vectors for Chinese Character Representations[C]//Processdings of NeurlPS.Vancouver CANADA:Neural Information Processing Systems,2019.
• [16] RAFFEL C,ELLIS DPW.Feed-Forward Networks with Attention can Solve Some Long-Term Memory Problems[J/OL].(2015-12-29).https://doi.org/10.48550/arXiv.1512.08756.
• [17] QU J F,DANTONG OUYANG,HUA W,et al.Distant supervision for neural relation extraction integrated with word attention and property features[J].Neural Networks.2018,100:59-69.
• [18] LI Z N,LI Q,ZOU X T,et al.Causality extraction based on self-attentive BiLSTM-CRF with transferred embeddings[J].Neurocomputing.2021,423(29):207-219.
• [19] LI S,ZHAO Z,HU R F,et al.Analogical Reasoning on Chinese Morphological and Semantic Relations[C]//Processdings of ACL.Melbourne,Australia:Association for Computational Linguistics,2018:138-143.
• [20] ABHISHEK KUMAR,DAISUKE KAWAHARA,SADAO KUROHASHI.Knowledge-enriched Two-layered Attention Network for Sentiment Analysis[C]//Processdings of NAACL.New Orleans,Louisiana:Association for Computational Linguistics,2018:253-258.