English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Multimodal fake news detection has attracted many research interests insocial forensics. Many existing approaches introduce tailored attentionmechanisms to guide the fusion of unimodal features. However, how thesimilarity of these features is calculated and how it will affect thedecision-making process in FND are still open questions. Besides, the potentialof pretrained multi-modal feature learning models in fake news detection hasnot been well exploited. This paper proposes a FND-CLIP framework, i.e., amultimodal Fake News Detection network based on Contrastive Language-ImagePretraining (CLIP). Given a targeted multimodal news, we extract the deeprepresentations from the image and text using a ResNet-based encoder, aBERT-based encoder and two pair-wise CLIP encoders. The multimodal feature is aconcatenation of the CLIP-generated features weighted by the standardizedcross-modal similarity of the two modalities. The extracted features arefurther processed for redundancy reduction before feeding them into the finalclassifier. We introduce a modality-wise attention module to adaptivelyreweight and aggregate the features. We have conducted extensive experiments ontypical fake news datasets. The results indicate that the proposed frameworkhas a better capability in mining crucial features for fake news detection. Theproposed FND-CLIP can achieve better performances than previous works, i.e.,0.7\%, 6.8\% and 1.3\% improvements in overall accuracy on Weibo, Politifactand Gossipcop, respectively. Besides, we justify that CLIP-based learning canallow better flexibility on multimodal feature selection.

Multimodal Fake News Detection via CLIP-Guided Learning