基于多模态的新能源汽车用户需求挖掘  

Multimodal-Based User Demand Mining for New Energy Vehicles

作　　者：陈佳佳 赵敬华[1] Jiajia Chen;Jinghua Zhao(Business School,University of Shanghai for Science and Technology,Shanghai)

机构地区：[1]上海理工大学管理学院,上海

出　　处：《建模与仿真》2025年第2期543-552,共10页Modeling and Simulation

基　　金：国家自然科学基金资助项目(72201173)。

摘　　要：本文针对新能源汽车市场快速扩张背景下,消费者在线评论中多模态信息的情感分析需求,提出了一种基于改进多头注意力机制的多模态情感分析模型。该模型通过跨模态和自注意力机制的融合,有效提升了新能源汽车在线评论中情感倾向的识别精度。实验结果表明,该模型在多个数据集上的性能优于现有方法,为新能源汽车用户需求挖掘提供了新的视角和工具。This paper addresses the need for sentiment analysis of multimodal consumer reviews in the rapidly expanding new energy vehicle market by proposing an improved multi-head attention mechanism-based multimodal sentiment analysis model.The model effectively enhances the recognition accuracy of emotional tendencies in online reviews by integrating cross-modality and self-attention mechanisms.Experimental results demonstrate that the model outperforms existing methods across multiple datasets,providing a new perspective and tool for mining user requirements in new energy vehicles.

关 键 词：新能源汽车 多模态 在线评论 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]