Different Absorption from the Same Sharing: Sifted Multi-task Learning for Fake News Detection
This work improves fake news detection for social media and information verification by incrementally enhancing multi-task learning to filter out adverse shared features.
The paper tackles the problem of fake news detection by addressing the issue of irrelevant or harmful shared features in multi-task learning, proposing a sifted multi-task learning method with a selected sharing layer that filters features using gate and attention mechanisms. The result is state-of-the-art performance on RumourEval and PHEME datasets, boosting F1-scores by over 0.87% and 1.31%, respectively.
Recently, neural networks based on multi-task learning have achieved promising performance on fake news detection, which focus on learning shared features among tasks as complementary features to serve different tasks. However, in most of the existing approaches, the shared features are completely assigned to different tasks without selection, which may lead to some useless and even adverse features integrated into specific tasks. In this paper, we design a sifted multi-task learning method with a selected sharing layer for fake news detection. The selected sharing layer adopts gate mechanism and attention mechanism to filter and select shared feature flows between tasks. Experiments on two public and widely used competition datasets, i.e. RumourEval and PHEME, demonstrate that our proposed method achieves the state-of-the-art performance and boosts the F1-score by more than 0.87%, 1.31%, respectively.