A veracity dissemination consistency-based few-shot fake news detection framework by synergizing adversarial and contrastive self-supervised learning.

In response to the aforementioned learning deficiency of conventional fully supervised PLM fine-tuning and the lack of annotated data for misinformation detection, there is an increasing trend towards fake news detection with minimal or no labeled data 7 , 10 , 11 .

also in 63,971 other papers