《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (3): 715-721.DOI: 10.11772/j.issn.1001-9081.2023030358

• 人工智能 • 上一篇    下一篇

基于注意力机制的多特征融合对话行为层次化分类方法

贾宗泽, 高鹏飞, 马应龙(), 刘晓峰, 夏海鑫   

  1. 华北电力大学 控制与计算机工程学院,北京 102206
  • 收稿日期:2023-04-04 修回日期:2023-06-06 接受日期:2023-06-08 发布日期:2023-07-04 出版日期:2024-03-10
  • 通讯作者: 马应龙
  • 作者简介:贾宗泽(1997—),男,山西运城人,硕士研究生,主要研究方向:自然语言处理
    高鹏飞(1996—),男,山东济宁人,硕士研究生,主要研究方向:机器学习
    刘晓峰(1994—),男,山西大同人,博士研究生,主要研究方向:自然语言处理
    夏海鑫(1997—),男,河北张家口人,硕士研究生,主要研究方向:知识图谱。
  • 基金资助:
    国家电网科技部项目(SGGSXT00XMJS2250023)

Multi-feature fusion attention-based hierarchical classification method for dialogue act

Zongze JIA, Pengfei GAO, Yinglong MA(), Xiaofeng LIU, Haixin XIA   

  1. School of Control and Computer Engineering,North China Electric Power University,Beijing 102206,China
  • Received:2023-04-04 Revised:2023-06-06 Accepted:2023-06-08 Online:2023-07-04 Published:2024-03-10
  • Contact: Yinglong MA
  • About author:JIA Zongze, born in 1997, M. S. candidate. His research interests include natural language processing.
    GAO Pengfei, born in 1996, M. S. candidate. His research interests include machine learning.
    LIU Xiaofeng, born in 1994, Ph. D. candidate. His research interests include natural language processing.
    XIA Haixin, born in 1997, M. S. candidate. His research interests include knowledge graph.
  • Supported by:
    Project of Science and Technology Department of State Grid(SGGSXT00XMJS2250023)

摘要:

目前深度学习模型在对话行为识别中被广泛采用,通过挖掘多种对话行为特征以提升对话行为分类性能。然而,这些方法忽视了不同对话行为特征之间的潜在关联和相互影响,且对话行为分类过程中也很少考虑对话行为标签之间的语义关联关系,这些都妨碍了对话行为识别的性能提升。针对以上问题,提出一种基于注意力机制的多特征融合层次化分类(MFA-HC)方法用于对话行为识别。首先,提出一种基于无遗忘学习的对话行为层次化分类框架,结合词、词性以及相关语言学统计量等多种细粒度特征来学习训练对话行为分类模型;其次,提出一种基于注意力机制的共性-个性模型捕获不同特征之间的共性和个性特征。在两个基准数据集SwDA(Switchboard Dialogue Act corpus)和MRDA(ICSI Meeting Recorder Dialogue Act corpus)上的实验结果表明:相较于目前整体性能较优的DARER(Dual-tAsk temporal Relational rEcurrent Reasoning network),MFA-HC方法通过捕捉话语中隐含的共性和个性特征,分类准确率分别提高了0.6%和0.1%。

关键词: 对话行为, 特征表示, 特征融合, 多特征, 层次分类

Abstract:

Nowadays, deep learning models have been widely applied in dialogue act recognition, which can improve classification performance by mining various features of dialogue acts. However, the existing methods neglect the latent association and interaction between different features of dialogue acts and also seldom consider the semantic relevance between labels of dialogue act in the classification process, which hinders from improving the performance of dialogue act recognition. To solve these problems, an MFA-HC (Multi-feature Fusion Attention-based Hierarchical Classification) method for recognizing dialogue act was proposed. Firstly, a hierarchical dialogue act classification framework based on learning without forgetting was proposed, which combined various fine-grained features such as words, parts of speech and relevant linguistic statistics to learn and train the dialogue act classification model. Secondly, a universality-individuality model based on attention mechanism was proposed to capture the universality and individuality features among different features. Experimental results on two benchmark datasets SwDA (Switchboard Dialogue Act corpus) and MRDA (ICSI Meeting Recorder Dialogue Act corpus) show that, compared with DARER (Dual-tAsk temporal Relational rEcurrent Reasoning network), which has the current overall superior performance in existing methods, MFA-HC method improves the classification accuracy by 0.6% and 0.1% by capturing the universality and individuality features hidden in the utterance.

Key words: dialogue act, feature representation, feature fusion, multi-feature, hierarchical classification

中图分类号: