A Multi-modal Rumor Detection Model Based on Temporal Graph Attention Network

DOI

10.2991/978-94-6463-734-2_99How to use a DOI?

Keywords

Rumor detection; Multi-modal; Graph attention network

Abstract

Objective: To address the issue of insufficient mining of structural and temporal sequence features of information dissemination in existing rumor detection methods, a multi-modal rumor detection model based on temporal graph attention is designed. Methods: For the text modality, a RoBERTa pre-trained model is used as the basis, and GAT and GRU modules are introduced to extract and fuse mixed features of text and dissemination structure. For the image modality, ViT is used to extract image features. Multi-modal features are fused through self-attention and cross-attention mechanisms to complete rumor detection. Results: The accuracy and F1 value of the proposed model on the Twitter dataset reach 91.1% and 91.4%, respectively, achieving the best performance in the comparative experiments. Limitations: The performance of the model on other datasets has not been tested. Conclusion: The proposed model can effectively improve the rumor detection effect of multi-modal posts on social media.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Shiming Li
PY  - 2025
DA  - 2025/05/27
TI  - A Multi-modal Rumor Detection Model Based on Temporal Graph Attention Network
BT  - Proceedings of the 2025 10th International Conference on Social Sciences and Economic Development (ICSSED 2025)
PB  - Atlantis Press
SP  - 890
EP  - 905
SN  - 2352-5428
UR  - https://doi.org/10.2991/978-94-6463-734-2_99
DO  - 10.2991/978-94-6463-734-2_99
ID  - Li2025
ER  -