广西师范大学学报(自然科学版) ›› 2019, Vol. 37 ›› Issue (1): 50-61.doi: 10.16088/j.issn.1001-6600.2019.01.006

• 第二十四届全国信息检索学术会议专栏 • 上一篇    下一篇

基于多任务深度学习的文本情感原因分析

余传明1, 李浩男2, 安璐3*   

  1. 1. 中南财经政法大学信息与安全工程学院,湖北武汉430074;
    2. 中南财经政法大学统计与数学学院,湖北武汉430074;
    3. 武汉大学信息管理学院,湖北武汉430072
  • 收稿日期:2018-10-14 出版日期:2019-01-20 发布日期:2019-01-08
  • 通讯作者: 安璐(1979—), 女, 湖北武汉人, 武汉大学教授,博导。 E-mail:anlu97@163.com
  • 基金资助:
    国家自然科学基金(71373286, 71603189);教育部哲学社会科学研究重大课题攻关项目(17JZD034)

Analysis of Text Emotion Cause Based on Multi-task Deep Learning

YU Chuanming1,LI Haonan2,AN Lu3*   

  1. 1. School of Information and Security Engineering, Zhongnan University of Economics and Law, Wuhan Hubei 430073, China;
    2. School of Statistics and Mathematics, Zhongnan University of Economics and Law, Wuhan Hubei 430073, China;
    3. School of Information Management, Wuhan University, Wuhan Hubei 430072, China
  • Received:2018-10-14 Online:2019-01-20 Published:2019-01-08

摘要: 多任务学习利用不同任务之间的相似性辅助决策,与单任务学习相比,多任务学习能够利用更多的信息,从而可以弥补单任务学习信息利用不足的缺陷。本文选择NTCIR-ECA数据集中的中文和英文文本数据作为实验数据,以情感原因分析作为研究任务,提出了一种结合多任务学习和深度学习的模型 MTDLM(multi-task deep learning model),实现不同语种下的情感原因分析。实验结果表明,在数据不平衡的情况下,MTDLM模型对英文语种的情感原因识别的最优F值为39%,优于单任务学习(F值为0)和传统基线模型(LR的F值为33%),从而验证了模型的有效性。

关键词: 情感原因分析, 多任务学习, 深度学习, 文本挖掘

Abstract: Multi-task learning utilizes the similarity between different tasks to help decision making. Compared with single-task learning, multi-task learning can use more information, which can make up for the deficiency of single-task learning in the use of information. In this paper, a NTCIR-ECA dataset, which contains Chinese and English text data is used as the date in the experiment. The emotional cause analysis is regarded as the research task and a multi-task deep learning model (MTDLM) which combines multi-task learning and deep learning is presented. Finally, this model is used to do the emotional cause analysis in different languages. The experimental results show that in the case of unbalanced data, the optimal F value of the MTDLM model for English language emotion recognition is 39%, superior to single task learning (F value is 0) and traditional baseline model (LR, F value is 33%). The validity of the model is thus verified.

Key words: emotion cause analysis, multi-task learning, deep learning, text-mining

中图分类号: 

  • TP391
[1] 袁丽. 基于文本的情绪自动归因方法研究[D].深圳: 哈尔滨工业大学深圳研究生院, 2014.
[2] 李然, 林政, 林海伦,等. 文本情绪分析综述[J]. 计算机研究与发展, 2018, 55(1):30-52.
[3] YANG C, LIN H Y, CHEN H H. Emotion classification using web blog corpora[C]//International Conference on Web Intelligence. Washington, DC:IEEE Computer Society, 2007:275-278.
[4] RAO Y. Contextual sentiment topic model for adaptive social emotion classification[J]. IEEE Intelligent Systems, 2016, 31(1):41-47.
[5] BANDHAKAVI A, WIRATUNGA N, PADMANABHAN D, et al. Lexicon based feature extraction for emotion text classification[J]. Pattern Recognition Letters, 2016,93(1):133-142.
[6] KWON H J, HONG K S. A framework of human emotion prediction based on a multi-dimensional emotion model[C]//Proceedings of the 7th International Conference on Information Security and Assurance. [S.l.]:SERSC, 2013: 171-174.
[7] LIU H H, LI S S, ZHOU G D, et al. Joint modeling of news reader’s and comment writer’s emotions[C]//Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics. Stroudsburg, ACL, 2013: 511-515.
[8] LEE S Y M, CHEN Y, HUANG C R, et al. Detecting emotion cause with a linguistic rule-based approach[J]. Computational Intelligence, 2013, 29(3):390-416.
[9] GAO K, XU H, WANG J. A rule-based approach to emotion cause detection for Chinese micro-blogs[J].Expert Systems with Application, 2015, 42(9):4517-4528.
[10] GUI L, YUAN L, XU R, et al. Emotion cause detection with linguistic construction in Chinese Weibo text[M]. Natural Language Processing and Chinese Computing. Berlin: Springer, 2014, 496 (s3/4):457-464
[11] XU R, HU J, LU Q, et al. An ensemble approach for emotion cause detection with event extraction and multi-kernel SVMs[J]. Tsinghua Science and Technology, 2017, 22(6): 646-659.
[12] CHEN Y, LEE S Y M, LI S, et al. Emotion cause detection with Linguistic constructions[C]//Proceedings of the 23rd International Conference on Computational Linguistics. Stroudsburg, ACL, 2010:179-187.
[13] 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.
[14] 梁斌, 刘全, 徐进,等. 基于多注意力卷积神经网络的特定目标情感分析[J]. 计算机研究与发展, 2017, 54(8):1724-1735.
[15] 何炎祥,孙松涛,牛菲菲, 等. 用于微博情感分析的一种情感语义增强的深度学习模型[J].计算机学报,2017,40(4):773-790.
[16] MOESKOPS P, WOLTERINK J M, VELDEN B H M V D, et al. Deep learning for multi-task medical image segmentation in multiple modalities[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham, Switzertand: Springer, 2016:478-486.
[17] FOURURE D, EMONET R, FROMONT E, et al. Multi-task, multi-domain learning: application to semantic segmentation and pose regression[J]. Neurocomputing, 2017, 251(1): 68-80.
[18] BANSAL T, BELANGER D, McCALLUM A. Ask the GRU:Multi-task learning for deep text recommendations[C]//Proceedings of the 10th ACM Conference on Recommender Systems. New York:ACM, 2016: 107-114.
[19] TIAN Y, WANG H, WANG X. Object localization via evaluation multi-task learning[J]. Neurocomputing, 2017, 253(1):34-41.
[20] YU B, LANE I. Multi-task deep learning for image understanding[C]//Soft Computing and Pattern Recognition. Washington DC:IEEE, 2015:37-42.
[21] 王晓栋, 严菲, 洪朝群. 一种基于半监督多任务学习的特征选择模型[J]. 厦门大学学报(自然科学版), 2017, 56(4):567-575.
[22] 范正光,屈丹,闫红刚,等. 基于深层神经网络的多特征关联声学建模方法[J]. 计算机研究与发展,2017,54(5):1036-1044.
[23] CHEN T, GUESTRIN C. XGBoost: a scalable tree boosting system[C]//ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2016:785-794.
[1] 张明宇, 赵猛, 蔡夫鸿, 梁钰, 王鑫红. 基于深度学习的波浪能发电功率预测[J]. 广西师范大学学报(自然科学版), 2020, 38(3): 25-32.
[2] 李维勇, 柳斌, 张伟, 陈云芳. 一种基于深度学习的中文生成式自动摘要方法[J]. 广西师范大学学报(自然科学版), 2020, 38(2): 51-63.
[3] 刘英璇, 伍锡如, 雪刚刚. 基于深度学习的道路交通标志多目标实时检测[J]. 广西师范大学学报(自然科学版), 2020, 38(2): 96-106.
[4] 张金磊, 罗玉玲, 付强. 基于门控循环单元神经网络的金融时间序列预测[J]. 广西师范大学学报(自然科学版), 2019, 37(2): 82-89.
[5] 黄丽明, 陈维政, 闫宏飞, 陈翀. 基于循环神经网络和深度学习的股票预测方法[J]. 广西师范大学学报(自然科学版), 2019, 37(1): 13-22.
[6] 岳天驰, 张绍武, 杨亮, 林鸿飞, 于凯. 基于两阶段注意力机制的立场检测方法[J]. 广西师范大学学报(自然科学版), 2019, 37(1): 42-49.
[7] 王祺, 邱家辉, 阮彤, 高大启, 高炬. 基于循环胶囊网络的临床语义关系识别研究[J]. 广西师范大学学报(自然科学版), 2019, 37(1): 80-88.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 广西师范大学学报(自然科学版)编辑部
地址:广西桂林市三里店育才路15号 邮编:541004
电话:0773-5857325 E-mail: gxsdzkb@mailbox.gxnu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发