Multi-view Fake News Detection method based on graph neural network

The rapid growth of content on social media-ranging from videos and images to text-has made information highly accessible but also increasingly vulnerable to manipulation.Malicious entities exploit this openness to spread false news that influences public opinion, especially during critical events.This phenomenon has driven significant attention toward automatic fake news detection.Our proposed approach addresses this challenge by integrating structural, temporal, and semantic cues within a dynamic graph-based framework.Unlike traditional methods that rely solely on static features or standalone natural language processing, our model captures the evolving nature of social interactions and news propagation patterns over time.This enables a more comprehensive and context aware detection of fake news, outperforming existing static graph models in both accuracy and temporal sensitivity.Therefore, in this paper, we propose a novel method called Multi-view Fake News Detection (MvFD), which simultaneously captures the temporal interaction of entities in social networks, user preferences, and news content, and utilizes the captured signals for modeling.Experimental results demonstrate the effectiveness of the model on real-world datasets.