An Attentive Convolutional Recurrent Network for Fake News Detection

Abstract

With the rapid growth of social media and online news platforms, the spread of fake news has become a major problem, leading to misinformation and distrust. In this paper, we propose an attentive convolutional recurrent network (ACRN) for fake news detection, which combines convolutional learning and recurrent learning capabilities to capture both local and global temporal information. Additionally, we incorporate attention mechanisms to focus on important features and reduce noise. We evaluate our model on a publicly available dataset and compare it with state-of-the-art methods. The results show that our ACRN model outperforms the existing methods in terms of accuracy, precision, recall, and F1-score. We also perform an ablation study to demonstrate the effectiveness of our attention mechanisms. Our proposed ACRN model can applied as a reliable computation intelligence tool for detecting fake news and improving the accuracy of news verification.