EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

TeFN: Text‐Centered Fusion Network With Multilevel Attention for Multimodal Sentiment Analysis

Wanzhi Chen, Yijia Liu, Yunhe Xiang, Yue Hou

International Journal of Intelligent Systems · 2026-01 · 2026 (1)

DOI: 10.1155/int/8485020 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Intelligent Systems, online erschienen am 29.05.2026. Verfasst von Wanzhi Chen und 3 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Science Foundation of Liaoning Province, Liaoning Technical University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the proliferation of social videos and voice‐content platforms, multimodal sentiment analysis (MSA) has become critical for applications such as user engagement analytics and AI‐driven customer service. In response to the problems of insufficient multimodal feature extraction, semantic feature differences between modalities, and weak interactions in MSA, a text‐centered fusion network with multilevel attention was proposed. We design a deep temporal feature extraction network to extract deep contextual temporal features from each modality. In addition, we develop a text‐centered multilevel attention interaction network for deep information interaction guided by the text, including a text‐centered cross‐modal transformer and an attention‐on‐attention transformer. We validate our model on two public datasets, CMU‐MOSI and CMU‐MOSEI. The experimental results demonstrate that our model effectively outperforms most existing methods in terms of accuracy in sentiment analysis tasks.

Förderung

  • Natural Science Foundation of Liaoning Province
  • Liaoning Technical University

Zitieren

APA

Chen, W., Liu, Y., Xiang, Y., & Hou, Y. (2026). TeFN: Text‐Centered Fusion Network With Multilevel Attention for Multimodal Sentiment Analysis. International Journal of Intelligent Systems, 2026(1). https://doi.org/10.1155/int/8485020

Vancouver

Chen W, Liu Y, Xiang Y, Hou Y. TeFN: Text‐Centered Fusion Network With Multilevel Attention for Multimodal Sentiment Analysis. International Journal of Intelligent Systems. 2026;2026(1). doi:10.1155/int/8485020

BibTeX
@article{Chen2026,
  author = {Chen, Wanzhi and Liu, Yijia and Xiang, Yunhe and Hou, Yue},
  title = {TeFN: Text‐Centered Fusion Network With Multilevel Attention for Multimodal Sentiment Analysis},
  journal = {International Journal of Intelligent Systems},
  year = {2026},
  volume = {2026},
  number = {1},
  doi = {10.1155/int/8485020}
}
Kostenlos starten →