A multimodal fake news detection model based on crossmodal attention residual and multichannel convolutional neural networks

作者：

Highlights：

• A multimodal fake news detection model based on crossmodal attention residual network and multichannel convolutional neural network is proposed.

• Our model can fuse the relevant information between different modalities while keeping the unique properties for each modality and alleviate the influence of noise information which may be generated by crossmodal fusion.

• The proposed method outperforms state-of-the-art methods and learns more discriminable feature representations.

• We contribute a large scale multimodal fake news dataset from Weibo platform and will make it available to the public.

摘要

•A multimodal fake news detection model based on crossmodal attention residual network and multichannel convolutional neural network is proposed.•Our model can fuse the relevant information between different modalities while keeping the unique properties for each modality and alleviate the influence of noise information which may be generated by crossmodal fusion.•The proposed method outperforms state-of-the-art methods and learns more discriminable feature representations.•We contribute a large scale multimodal fake news dataset from Weibo platform and will make it available to the public.

论文关键词：Fake news detection,Crossmodal attention,Residual network,Convolutional neural network

论文评审过程：Received 29 February 2020, Revised 27 September 2020, Accepted 8 November 2020, Available online 16 November 2020, Version of Record 16 November 2020.

论文官网地址：https://doi.org/10.1016/j.ipm.2020.102437