基于深度学习的多模态情感分析研究综述  

Survey of Multimodal Sentiment Analysis Based on Deep Learning

作　　者：李文潇 梅红岩[1] 李雨恬 LI Wen-xiao;MEI Hong-yan;LI Yu-tian(School of Electronics&Information Engineering,Liaoning University of Technology,Jinzhou 121001,China)

机构地区：[1]辽宁工业大学电子与信息工程学院,辽宁锦州121001

出　　处：《辽宁工业大学学报（自然科学版）》2022年第5期293-298,共6页Journal of Liaoning University of Technology(Natural Science Edition)

摘　　要：对深度学习完成多模态情感分析进行了综述。简要介绍了多模态情感分析的概念、背景及意义;详细介绍了深度学习在多模态情感分析当中的应用;总结了多模态融合技术和交互技术;讨论了多模态情感分析未来发展的走向。This paper mainly reviews the multimodal sentiment analysis completed by deep learning:first,the concept,background and significance of multimodal sentiment analysis are briefly introduced;secondly,the application of deep learning in multimodal sentiment analysis is introduced in detail;then the multimodal fusion is summarized.The multi-modal fusion technology and interaction technology are summarized.The future development trend of multimodal sentiment analysis is discussed.

关 键 词：深度学习 情感分析 多模态 模态融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]