Bootstrapping Multi-view Representations for Fake News Detection
This addresses the problem of detecting fake news in multimedia content for social media and information verification, representing an incremental improvement with novel feature fusion techniques.
The paper tackles fake news detection by proposing a Bootstrapping Multi-view Representations (BMR) scheme that extracts and fuses text, image pattern, and image semantics views, using improved Multi-gate Mixture-of-Expert networks and bootstrapping to outperform state-of-the-art methods in experiments on typical datasets.
Previous researches on multimedia fake news detection include a series of complex feature extraction and fusion networks to gather useful information from the news. However, how cross-modal consistency relates to the fidelity of news and how features from different modalities affect the decision-making are still open questions. This paper presents a novel scheme of Bootstrapping Multi-view Representations (BMR) for fake news detection. Given a multi-modal news, we extract representations respectively from the views of the text, the image pattern and the image semantics. Improved Multi-gate Mixture-of-Expert networks (iMMoE) are proposed for feature refinement and fusion. Representations from each view are separately used to coarsely predict the fidelity of the whole news, and the multimodal representations are able to predict the cross-modal consistency. With the prediction scores, we reweigh each view of the representations and bootstrap them for fake news detection. Extensive experiments conducted on typical fake news detection datasets prove that the proposed BMR outperforms state-of-the-art schemes.