计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (11A): 221100065-6.doi: 10.11896/jsjkx.221100065

• 人工智能 • 上一篇    下一篇

基于特征融合的无标复句关系识别

杨进才1, 马晨1, 肖明2   

  1. 1 华中师范大学计算机学院 武汉 430079
    2 华中师范大学语言与语言教育研究中心 武汉 430079
  • 发布日期:2023-11-09
  • 通讯作者: 杨进才(jcyang@mail.ccnu.edu.cn)
  • 基金资助:
    国家社科基金(19BYY092)

Relation Recognition of Unmarked Complex Sentences Based on Feature Fusion

YANG Jincai1, MA Chen1, XIAO Ming2   

  1. 1 School of Computer Science,Central China Normal University,Wuhan 430079,China
    2 Research Center for Language and Language Education,Central China Normal University,Wuhan 430079,China
  • Published:2023-11-09
  • About author:YANG Jincai,born in 1976,doctor,professor,doctoral supervisor,is a member of China Computer Federation.His main research interests include advanced database and information system,Chinese information processing,artificial intelligence and natural language processing.
  • Supported by:
    National Social Science Fund of China(19BYY092).

PDF (PC)

487

摘要： 无标复句因缺少关联词的辅助,其关系识别为自然语言处理中的一项较为困难的任务。将词性特征融入到词向量中,训练得到含有外部特征的词向量表示,通过组合BERT模型与BiLSTM模型,将字向量、词向量、词性向量结合进行训练,并在特征融合层添加BiLSTM模型捕获的极性特征信息以及CNN模型捕获的依存句法特征信息。实验结果表明,该方法在汉语复句分类上取得了较好的效果,与基准模型相比在宏F1值与微F1值上均有提升,在顶层分类上取得了83.67%的微F1值,在第二层分类上取得了68.28%的微F1值。

关键词: 无标复句, BERT, 特征融合, 深度学习

Abstract: Unlike marked complex sentences,which lack the assistance of relation words,the identification of unmarked complex sentences is a difficult task in natural language processing.Integrating part of speech features into word vectors,and the word vector representation containing external features is obtained by training.By combining the BERT model and the BiLSTM model,the word vector and the part-of-speech vector are combined for training,and the polar feature information captured by BiLSTM model and the dependency syntax feature information captured by CNN model are added to the feature fusion layer.Experimental results show that the methods of adding features and combining multiple deep learning models can achieve better results in classification of Chinese complex sentences.Compared with the benchmark model,the macro F1 value and micro F1 value are improved.The best classification effect achieves 83.67% micro F1 value in the top layer classification and 68.28% micro F1 value in the second layer classification.

Key words: Unmarked complex sentence, BERT, Feature fusion, Deep learning

中图分类号: 

  • TP391
[1]XING F Y.Research on Chinese Complex Sentences[M].Beijing:The Commercial Press,2001:1-37.
[2]COHAN A,DERNONCOURT F,KIM D S,et al.A Discourse-Aware AttentionModel for Abstractive Summarization of Long Documents[C]//Proceedings of NAACL.2018:615-621.
[3]TANG J,LIN H,LIAO M,et al.From Discourse to Narrative:KnowledgeProjection for Event Relation Extraction[C]//Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing(Volume 1:Long Papers).2021:732-742.
[4]LIU Y,LI S,ZHANG X,et al.Implicit Discourse Relation Classification via Multi-Task Neural Networks[J].AAAI Press,2016,30(1):2750-2756.
[5]SUN J,LI Y C,ZHOU G D,et al.Research of Chinese Implicit Discourse Relation Recognition[J].Acta Scientiarum Naturalium Universitatis Pekinensis,2014,50(1):111-117.
[6]KIM Y.Convolutional Neural Networks for Sentence Classification[C]//Conference on Empirical Methods in Natural Language Processing(EMNLP 2014).Doha,Qatar:Association for Computational Linguistics,2014:1746-1751.
[7]ZENG D J,LIU K,LAI S W,et al.Relation Classification viaConvolutional Deep Neural Network[C]//The 25th International Conference on Computational Linguistics:Technical Papers(COLING 2014).2014:2335-2344.
[8]CAI R,ZHANG X D,WANG H F.Bi-directional recurrent convolutional neural network for relation classification[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics(Volume 1:Long Papers).2016:756-765.
[9]FAN Z W,ZHANG M,LI Z H.BiLSTM-based Implicit Dis-course Relation Classification Combining Self-attentionMechanism and Syntactic Information[J].Journal of Computer Science,2019,46(5):214-220.
[10]RÖNNQVIST S,SCHENK N,CHIARCOS C.A RecurrentNeural Model with Attention for the Recognition of Chinese Implicit Discourse Relations[C]//Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics(Volume 2:Short Papers).2017:256-262.
[11]TIAN W H,GAO Y Q,HUANG H W,et al.Implicit DiscourseRelation Analysis Based on Multi-task Bi-LSTM[J].Journal of Chinese information processing,2019,33(5):47-53.
[12]JI J H,ZHANG M Y,QIN B,et al.The Chinese Discourse Parser[J].Journal of Jiangxi Normal University(Natural Science),2015(2):124-131.
[13]WANG T S,LI P F,ZHU Q M.Chinese Implicit Discourse Relation Recognition Based on Data Augmentation[J].Journal of Computer Science,2021,48(10):85-90.
[14]TIAN J,ZHU D J,LONG H.Chinese Short Text Multi-Classifi-
cation Based on Word and Part-of-Speech Tagging Embedding[C]//Proceedings of the 2018 International Conference on Algorithms,Computing and Artificial Intelligence,2018:1-6.
[15]KALARANI P,SELVA B S.Sentiment Analysis by POS and JointSentiment Topic Features Using SVM and ANN[J].SoftComputing,2019,23(16):7067-7079.
[16]HEH Y,ZHENG J,ZHANG Z P.Text Sentiment AnalysisCombined with Part of Speech Features and ConvolutionalNeural Network[J].Computer Engineering,2018,44(11):209-214,221.
[17]ZHANG X H,LIU L X,DAI X Y,et al.CNN_BIGRU TextClassification Model Based On Part of Speech Features[J].Computer Applications and Software,2021,38(11):155-161.
[18]SHEN Y,LIN Z,HUANG C W,et al.Neural Language Modeling by Jointly Learning Syntax and Lexicon[C]//International Conference on Learning Representations.2018.
[19]SHEN Y,TAN S,SORDONI A,et al.Ordered Neurons:In-tegrating Tree Structures into Recurrent Neural Networks[C]//International Conference on Learning Representations.2019.
[20]PENG H,SCHWARTZ R,SMITH N A.PaLM:A Hybrid Parser and Language Model[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing(EMNLP-IJCNLP).2019:3644-3651.
[21]SUN K L,DENG D H,LI Y,et al.Inner-Attention Based Multi-Way Convolutional Neural Network for Relation Recognition in Chinese Compound Sentence[J].Journal of Chinese information processing,2020,34(6):9-17,26.
[22]YANG J C,WANG Y Y,CAO Y,et al.Relation Classification of Non-Saturated Chinese Compound Sentence via Feature Fusion CNN[J].Computer Systems and Applications,2020,29(6):224-229.
[23]LI S,KONG F,ZHOU G D.Recognizing PDTB Style Implicit Discourse Relations[J].Journal of Computer Science,2016,30(4):81-89.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发