基于AM-BiLSTM模型的情绪原因识别
|
摘要
文本情绪原因识别是情感分析中一个新的研究方向,旨在从文本中自动检测出导致某一情绪产生的原因。针对循环神经网络在长文中出现的长期依赖问题,本文提出了一种基于注意力机制和双向长短时记忆(attention model and bi-directional long short-term memory,AM-BiLSTM)神经网络模型的情绪原因识别方法。该方法采用字符向量表示文本语义信息,使用BiLSTM模型提取文本特征,该过程结合了人工提取的子句特征,在训练模型时,引入了注意力机制来优化模型性能,使用softmax对子句进行分类。实验结果表明本文方法对情绪原因的识别是有效的。
Textual emotion cause recognition is a new research direction in sentiment analysis, which aims to identify the causes of a certain emotion from the natural language texts automatically. In order to solve the long-term dependency problem of recurrent neural network in long natural language texts, this paper proposes an emotion cause recognition model based on attention mechanism(AM) and bi-directional long short-term memory(BiLSTM). The AM-BiLSTM model adopts character embedding to represent semantic information of the text, and then the BiLSTM is used to extract the textual features. In the model training phase, the AM is introduced to optimize the performance, and finally, we use softmax to classify the text. The experimental results show that our model is effective in the emotion cause recognition task.
Textual emotion cause recognition is a new research direction in sentiment analysis, which aims to identify the causes of a certain emotion from the natural language texts automatically. In order to solve the long-term dependency problem of recurrent neural network in long natural language texts, this paper proposes an emotion cause recognition model based on attention mechanism(AM) and bi-directional long short-term memory(BiLSTM). The AM-BiLSTM model adopts character embedding to represent semantic information of the text, and then the BiLSTM is used to extract the textual features. In the model training phase, the AM is introduced to optimize the performance, and finally, we use softmax to classify the text. The experimental results show that our model is effective in the emotion cause recognition task.
引文
[1]EKMAN P.Expression and the nature of emotion[J].Approaches to Emotion,1984,3:324-344.
[2]雷龙艳.中文微博细粒度情绪识别研究[D].衡阳:南华大学,2014.LEI L Y.Research on Fine-grained Sentiment Analysis Base on Chinese Micro-Blog[D].Hengyang:University of South China,2014(Ch).
[3]LEE S.A Linguistic Approach to Emotion Detection and Classification[D].Hong Kong:Hong Kong Polytechnic University,2010.
[4]何跃,邓唯茹,张丹.中文微博的情绪识别与分类研究[J].情报杂志,2014(2):136-139.HE Y,DENG W R,ZHANG D.Study on sentiments recognition and classification of Chinese micro-blog[J].Journal of Information,2014(2):136-139.DOI:10.3969/j.issn.1002-1965.2014.02.026(Ch).
[5]朱少杰.基于深度学习的文本情感分类研究[D].哈尔滨:哈尔滨工业大学,2014.ZHU S J.Research on Text Sentiment Classification Based on Deep Learning[D].Harbin:Harbin Institute of Technology,2014(Ch).
[6]WANG Y Q,FENG S,WANG D L,et al.Multi-label Chinese microblog emotion classification via convolutional neural network[C]//Proceedings of Asia-Pacific Web Conference(LNCS 9931).Cham:Springer,2016:567-580.DOI:10.1007/978-3-319-45814-4_46.
[7]CHEN Y,LEE S Y M,LI S S,et al.Emotion cause detection with linguistic constructions[DB/OL].[2018-05-01].https://pdfs.semanticscholar.org/df12/d09c5f12ef8c6aceeda4f83a25f807e0ea45.pdf?_ga=2.87405656.1390102688.1555057393-1962158694.1555057393.
[8]袁丽.基于文本的情绪自动归因方法研究[D].哈尔滨:哈尔滨工业大学,2014.YUAN L.The Study on Text-Based Emotion Cause Detection[D].Harbin:Harbin Institute of Technology,2014(Ch).
[9]李逸薇,李寿山,黄居仁,等.基于序列标注模型的情绪原因识别方法[J].中文信息学报,2013,27(5):93-99.LI Y W,LI S S,HUANG J R,et al.Detecting emotion cause with sequence labeling model[J].Journal of Chinese Information Processing,2013,27(5):93-99.DOI:10.3969/j.issn.1003-0077.2013.05.013(Ch).
[10]慕永利,李旸,王素格,等.基于E-CNN的情绪原因识别方法[J].中文信息学报,2018,32(2):120-128.MU Y L,LI Y,WANG S G,et al.Emotion cause detection based on ensembled convolution neural netowrks[J].Journal of Chinese Information Processing,2018,32(2):120-128(Ch).
[11]刘飞龙,郝文宁,陈刚,等.基于双线性函数注意力Bi-LSTM模型的机器阅读理解[J].计算机科学,2017,44(S1):92-96.LIU F L,HAO W Y,CHENG G,et al.Attention of bilinear function based Bi-LSTM model for machine reading comprehension[J].Computer Science,2017,44(s1):92-96(Ch).
[12]MIKOLOV T,CHEN K,CORRADO G,et al.Efficient Estimation of Word Representations in Vector Space[DB/OL].[2018-03-02].https://arxiv.org/pdf/1301.3781.pdf.
[13]YANG Z C,YANG D Y,DYER C,et al.Hierarchical Attention Networks for Document Classification[DB/OL].[2018-04-06].https://www.researchgate.net/publication/305334401_Hierarchical_Attention_Networks_for_Document_Classification.DOI:10.18653/v1/N16-1174.
[14]GAO Q H,HU J N,XU R F,et al.Overview of NTCIR-13 ECA task[C]//Proceedings of the 13th NTCIR Conference on Evaluation of Information Access Technologies.Tokyo:National Institute of Informatics,2017:361-366.
[15]GUI L,YUAN L,XU R F,et al.Emotion cause detection with linguistic construction in Chinese Weibo text[C]//Proceedings of the 3rd CCF International Conference on Natural Language Processing and Chinese Computing.Berlin:Springer,2014:457-464.DOI:10.1007/978-3-662-45924-9_42.
[16]GUI L,WU D Y,XU R F,et al.Event-Driven Emotion Cause Extraction with Corpus Construction[DB/OL].[2018-05-01]https://www.aclweb.org/anthology/D16-1170.DOI:10.18653/v1/D16-1170.
[2]雷龙艳.中文微博细粒度情绪识别研究[D].衡阳:南华大学,2014.LEI L Y.Research on Fine-grained Sentiment Analysis Base on Chinese Micro-Blog[D].Hengyang:University of South China,2014(Ch).
[3]LEE S.A Linguistic Approach to Emotion Detection and Classification[D].Hong Kong:Hong Kong Polytechnic University,2010.
[4]何跃,邓唯茹,张丹.中文微博的情绪识别与分类研究[J].情报杂志,2014(2):136-139.HE Y,DENG W R,ZHANG D.Study on sentiments recognition and classification of Chinese micro-blog[J].Journal of Information,2014(2):136-139.DOI:10.3969/j.issn.1002-1965.2014.02.026(Ch).
[5]朱少杰.基于深度学习的文本情感分类研究[D].哈尔滨:哈尔滨工业大学,2014.ZHU S J.Research on Text Sentiment Classification Based on Deep Learning[D].Harbin:Harbin Institute of Technology,2014(Ch).
[6]WANG Y Q,FENG S,WANG D L,et al.Multi-label Chinese microblog emotion classification via convolutional neural network[C]//Proceedings of Asia-Pacific Web Conference(LNCS 9931).Cham:Springer,2016:567-580.DOI:10.1007/978-3-319-45814-4_46.
[7]CHEN Y,LEE S Y M,LI S S,et al.Emotion cause detection with linguistic constructions[DB/OL].[2018-05-01].https://pdfs.semanticscholar.org/df12/d09c5f12ef8c6aceeda4f83a25f807e0ea45.pdf?_ga=2.87405656.1390102688.1555057393-1962158694.1555057393.
[8]袁丽.基于文本的情绪自动归因方法研究[D].哈尔滨:哈尔滨工业大学,2014.YUAN L.The Study on Text-Based Emotion Cause Detection[D].Harbin:Harbin Institute of Technology,2014(Ch).
[9]李逸薇,李寿山,黄居仁,等.基于序列标注模型的情绪原因识别方法[J].中文信息学报,2013,27(5):93-99.LI Y W,LI S S,HUANG J R,et al.Detecting emotion cause with sequence labeling model[J].Journal of Chinese Information Processing,2013,27(5):93-99.DOI:10.3969/j.issn.1003-0077.2013.05.013(Ch).
[10]慕永利,李旸,王素格,等.基于E-CNN的情绪原因识别方法[J].中文信息学报,2018,32(2):120-128.MU Y L,LI Y,WANG S G,et al.Emotion cause detection based on ensembled convolution neural netowrks[J].Journal of Chinese Information Processing,2018,32(2):120-128(Ch).
[11]刘飞龙,郝文宁,陈刚,等.基于双线性函数注意力Bi-LSTM模型的机器阅读理解[J].计算机科学,2017,44(S1):92-96.LIU F L,HAO W Y,CHENG G,et al.Attention of bilinear function based Bi-LSTM model for machine reading comprehension[J].Computer Science,2017,44(s1):92-96(Ch).
[12]MIKOLOV T,CHEN K,CORRADO G,et al.Efficient Estimation of Word Representations in Vector Space[DB/OL].[2018-03-02].https://arxiv.org/pdf/1301.3781.pdf.
[13]YANG Z C,YANG D Y,DYER C,et al.Hierarchical Attention Networks for Document Classification[DB/OL].[2018-04-06].https://www.researchgate.net/publication/305334401_Hierarchical_Attention_Networks_for_Document_Classification.DOI:10.18653/v1/N16-1174.
[14]GAO Q H,HU J N,XU R F,et al.Overview of NTCIR-13 ECA task[C]//Proceedings of the 13th NTCIR Conference on Evaluation of Information Access Technologies.Tokyo:National Institute of Informatics,2017:361-366.
[15]GUI L,YUAN L,XU R F,et al.Emotion cause detection with linguistic construction in Chinese Weibo text[C]//Proceedings of the 3rd CCF International Conference on Natural Language Processing and Chinese Computing.Berlin:Springer,2014:457-464.DOI:10.1007/978-3-662-45924-9_42.
[16]GUI L,WU D Y,XU R F,et al.Event-Driven Emotion Cause Extraction with Corpus Construction[DB/OL].[2018-05-01]https://www.aclweb.org/anthology/D16-1170.DOI:10.18653/v1/D16-1170.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |