False News Detection Based on Adaptive Fusion of Multimodal Features

To address the challenge of underutilizing multimodal information in social media news for effective news detection, a false news detection model based on the adaptive fusion of multimodal features is proposed. Initially, semantic features of the news text, emotional features of the text, and semantic disparity features between pictures and text are individually extracted and represented. Subsequently, various features are weighted and fused by incorporating adaptive weight parameters to mitigate redundant information introduced during model concatenation. Finally, the fused features are input into the classifier to distinguish between true and false news. The actual test results demonstrate that the proposed model outperforms current state-of-the-art models in terms of the F1 value and other evaluation metrics. This model significantly enhances the performance of false news detection, providing robust support for detecting false news in social media.