Rumor Detection Model Based on GAT and Multimodal Feature Codingmes
Authors:
- 8 November 2024
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 799-805
Abstract
This paper proposes a social media rumor detection model that combines graph attention network (GAT) and multimodal feature coding. The model uses the multimodal content of social media news posts to encode, and on this basis, it also uses the graph attention network to construct the rumor propagation structure graph. At the same time, combined with the fact that rumor disseminators often publish social media rumors, it innovatively joins the multimodal feature mining of user historical posts, integrates the above work for common consideration, and carries out experiments on two commonly used data sets for rumor detection. The results show that the detection accuracy of the model is 7.72 and 2.14 percentage points higher than that of the commonly used rumor detection models Upfd and MFLAN, which proves the application potential of the multimodal fusion graph attention network rumor detection model in complex data environment.Keywords
This publication has 14 references indexed in Scilit:
- User Preference-aware Fake News Detection2021
- FakeBERT: Fake news detection in social media with a BERT-based deep learning approachMultimedia Tools and Applications, 2021
- EANN2018
- Leveraging Intra-User and Inter-User Representation Learning for Automated Hate Speech Detection2018
- Multimodal Fusion with Recurrent Neural Networks for Rumor Detection on Microblogs2017
- A Convolutional Approach for Misinformation Identification2017
- Deep Residual Learning for Image Recognition2016
- False rumors detection on Sina Weibo by propagation structures2015
- Prominent Features of Rumor Propagation in Online Social Media2013
- Information credibility on twitter2011