《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (11): 3603-3609.DOI: 10.11772/j.issn.1001-9081.2021091632

• 人工智能 • 上一篇    

基于迁移学习的文本共情预测

李晨光1, 张波2, 赵骞2, 陈小平1, 王行甫1()   

  1. 1.中国科学技术大学 计算机科学与技术学院,合肥 230026
    2.国网安徽省电力有限公司,合肥 230022
  • 收稿日期:2021-09-15 修回日期:2022-01-17 接受日期:2022-01-28 发布日期:2022-11-14 出版日期:2022-11-10
  • 通讯作者: 王行甫
  • 作者简介:李晨光(1999—),男,河南许昌人,硕士研究生,主要研究方向:情感识别、自然语言处理
    张波(1966—),男,安徽淮南人,高级工程师,硕士,主要研究方向:电力营销服务管理
    赵骞(1976—),男,安徽合肥人,高级工程师,硕士,主要研究方向:电力营销服务管理
    陈小平(1955—),男,重庆人,教授,博士,主要研究方向:智能体形式化建模、多机器人系统
    王行甫(1965—),男,安徽合肥人,副教授,博士,主要研究方向:自然语言处理、情感分析。cg0808@mail.ustc.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(92048301);安徽省电力有限公司科技项目(52120018004x)

Empathy prediction from texts based on transfer learning

Chenguang LI1, Bo ZHANG2, Qian ZHAO2, Xiaoping CHEN1, Xingfu WANG1()   

  1. 1.College of Computer Science and Technology,University of Science and Technology of China,Hefei Anhui 230026,China
    2.State Grid Anhui Electric Power Company Limited,Hefei Anhui 230022,China
  • Received:2021-09-15 Revised:2022-01-17 Accepted:2022-01-28 Online:2022-11-14 Published:2022-11-10
  • Contact: Xingfu WANG
  • About author:LI Chenguang, born in 1999, M. S. candidate. His research interests include emotion recognition, natural language processing.
    ZHANG Bo, born in 1966, M. S., senior engineer. His research interests include power marketing service management.
    ZHAO Qian, born in 1976, M. S., senior engineer. His research interests include power marketing service management.
    CHEN Xiaoping, born in 1955, Ph. D., professor. His research interests include agent formal modeling, multi‑robot system.
    WANG Xingfu, born in 1965, Ph. D., associate professor. His research interests include natural language processing, emotional analysis.
  • Supported by:
    National Natural Science Foundation of China(92048301);Science and Technology Project of Anhui Electric Power Company Limited(52120018004x)

摘要:

由于缺乏足够的训练数据,文本共情预测的进展一直都较为缓慢;而与之相关的文本情感极性分类任务则存在大量有标签的训练样本。由于文本共情预测与文本情感极性分类两个任务间存在较大相关性,因此提出了一种基于迁移学习的文本共情预测方法,该方法可从情感极性分类任务中学习到可迁移的公共特征,并通过学习到的公共特征辅助文本共情预测任务。首先通过一个注意力机制对两个任务间的公私有特征进行动态加权融合;其次为了消除两个任务间的数据集领域差异,通过一种对抗学习策略来区分两个任务间的领域独有特征与领域公共特征;最后提出了一种Hinge?loss约束策略,使共同特征对不同的目标标签具有通用性,而私有特征对不同的目标标签具有独有性。在两个基准数据集上的实验结果表明,相较于对比的迁移学习方法,所提方法的皮尔逊相关系数(PCC)和决定系数(R2)更高,均方误差(MSE)更小,充分说明了所提方法的有效性。

关键词: 迁移学习, 文本共情预测, 文本情感极性分类, 自然语言处理, 深度学习

Abstract:

Empathy prediction from texts achieves little progress due to the lack of sufficient labeled data, while the related task of text sentiment polarity classification has a large number of labeled samples. Since there is a strong correlation between empathy prediction and polarity classification, a transfer learning?based text empathy prediction method was proposed. Transferable public features were learned from the sentiment polarity classification task to assist text empathy prediction task. Firstly, a dynamic weighted fusion of public and private features between two tasks was performed through an attention mechanism. Secondly, in order to eliminate domain differences in datasets between two tasks, an adversarial learning strategy was used to distinguish the domain?unique features from the domain?public features between two tasks. Finally, a Hinge?loss constraint strategy was proposed to make common features be generic for different target labels and private features be unique to different target labels. Experimental results on two benchmark datasets show that compared to the comparison transfer learning methods, the proposed method has higher Pearson Correlation Coefficient (PCC) and coefficient of determination (R2), and has lower Mean?Square Error (MSE), which fully demonstrates the effectiveness of the proposed method.

Key words: transfer learning, text empathy prediction, text sentiment polarity classification, Nature Language Processing (NLP), deep learning

中图分类号: