情感分析是人工智能领域中重要的研究课题，在人工客服情绪安抚、抑郁症患者状态分类、刑侦辅助心理研究等多个方面都有着广泛使用。结合多个模态进行情感分析可以有效改善单模态情感分析的局限性，本文围绕多模态情感分析介绍了模态数据表示、融合方法、模型使用等方面，梳理概括相关研究内容。最后，总结了多模态情感分析存在的问题，对多模态情感分析的未来进行了展望和探讨。

Sentiment analysis is one of the major research topics in the field of artificial intelligence, and it is widely used in emotional comfort of artificial customer service, classification of depression patients, and research of criminal investigation. Combining multiple modalities in sentiment analysis can effectively eliminate the limitations of single‑modal. This paper introduces data representation, fusion methods, and model improvement around multi‑modal sentiment analysis, and summarizes related research content. Finally, it summarizes the existing problems of multi‑modal sentiment analysis, and discusses the future prospects of multi‑modal sentiment analysis.