See How You Read? Multi-Reading Habits Fusion Reasoning for Multi-Modal Fake News Detection

The existing approaches based on different neural networks automatically capture and fuse the multimodal semantics of news, which have achieved great success for fake news detection. However, they still suffer from limitations both shallow fusion features less attention to inconsistency between modalities. To overcome them, we propose multi-reading habits reasoning (MRHFR) multi-modal In MRHFR, inspired by people's reading summarize three basic cognitive put forward cognition-aware layer learn dependencies so as deepen their semantic-level integration. explore modalities develop coherence constraint two perspectives, first measures semantic consistency comments modal then probes deviation caused unimodal content through strategy. Experiments public datasets not only demonstrate that MRHFR achieves excellent performance but also provides a new paradigm capturing inconsistencies news.