Text-Guided Fine-grained Counterfactual Inference for Short Video Fake News Detection
Linlin Zong, Wenmin Lin, Jiahui Zhou, Xinyue Liu, Xianchao Zhang, Bo Xu, Shimin Wu
Abstract
Detecting fake news in short videos is crucial for combating misinformation. Existing methods utilize topic modeling and co-attention mechanism, overlooking the modality heterogeneity and resulting in suboptimal performance. To address this issue, we introduce Text-Guided Fine-grained Counterfactual Inference for Short Video Fake News detection (TGFC-SVFN). TGFC-SVFN leverages modality bias removal and teacher-model-enhanced inter-modal knowledge distillation to integrate the heterogeneous modalities in short videos. Specifically, we use causality-based reasoning prompts guided text as teacher model, which then transfers knowledge to the video and audio student models. Subsequently, a multi-head attention mechanism is employed to fuse information from different modalities. In each module, we utilize fine-grained counterfactual inference based on a diffusion model to eliminate modality bias. Experimental results on publicly available fake short video news datasets demonstrate that our method outperforms state-of-the-art techniques.
BibTeX
@article{Zong_Lin_Zhou_Liu_Zhang_Xu_Wu_2025, title={Text-Guided Fine-grained Counterfactual Inference for Short Video Fake News Detection}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/32112}, DOI={10.1609/aaai.v39i1.32112}, abstractNote={Detecting fake news in short videos is crucial for combating misinformation. Existing methods utilize topic modeling and co-attention mechanism, overlooking the modality heterogeneity and resulting in suboptimal performance. To address this issue, we introduce Text-Guided Fine-grained Counterfactual Inference for Short Video Fake News detection (TGFC-SVFN). TGFC-SVFN leverages modality bias removal and teacher-model-enhanced inter-modal knowledge distillation to integrate the heterogeneous modalities in short videos. Specifically, we use causality-based reasoning prompts guided text as teacher model, which then transfers knowledge to the video and audio student models. Subsequently, a multi-head attention mechanism is employed to fuse information from different modalities. In each module, we utilize fine-grained counterfactual inference based on a diffusion model to eliminate modality bias. Experimental results on publicly available fake short video news datasets demonstrate that our method outperforms state-of-the-art techniques.}, number={1}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Zong, Linlin and Lin, Wenmin and Zhou, Jiahui and Liu, Xinyue and Zhang, Xianchao and Xu, Bo and Wu, Shimin}, year={2025}, month={Apr.}, pages={1237-1245} }