Answers are grounded in this paper and cite their sources.
This work intends to analyze the multimodal emotion of online product reviews during e-commerce live broadcasts. To this end, the working principles of deep neural networks—such as long short-term memory, gated multimodal unit (GMU), and dynamic memory network (DMN)—are analyzed. Then, based on GMU and DMN, a DMN-GMU fusion model is built to conduct the multimodal emotion analysis of online product reviews. Finally, its performance is analyzed by comparing it with other models, using the online reviews of Zhongguancun online mobile phones as the experimental dataset. The results show that compared with other multimodal emotion analysis models, the classification accuracy and macro average F1-score of the DMN-GMU model are 3.08% and 3.31% higher, and the average classification time of each review pair is 0.29 ms longer.
No takes yet. Share an insight, caveat, or question.
Sun et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: