AUTHOR=Guo Ying , Ge Hong , Li Jinhong 

TITLE=A two-branch multimodal fake news detection model based on multimodal bilinear pooling and attention mechanism

JOURNAL=Frontiers in Computer Science

VOLUME=Volume 5 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/computer-science/articles/10.3389/fcomp.2023.1159063

DOI=10.3389/fcomp.2023.1159063

ISSN=2624-9898

ABSTRACT=The widespread fake news in various domains has a great negative impact on social life. Meanwhile, fake news with both textual and visual content is more convincing than textual content alone and spreads quickly on social media. Therefore, the detection of fake news is an urgent task for the current society. Concern the problem of insufficient feature extraction, and inability to fuse multimodality features effectively in fake news detection. In this paper we propose a detection method to detect fake news by fusing text and visual data. Firstly, we use two-branch to learn hidden layer information of modality to obtain more helpful features. Then we proposed multimodal bilinear pooling mechanism to better fuse textual and visual features and attention mechanism to capture single-modal internal relations for fake news detection. Experimental results demonstrated that our methodology outperforms the present state-of-the-art method on publicly available Weibo and Twitter dataset.