Research Article
Detection Method of Fake News Spread in Social Network Based on Deep Learning
@INPROCEEDINGS{10.1007/978-3-031-28867-8_35, author={Yandan Lu and Hongmei Ye}, title={Detection Method of Fake News Spread in Social Network Based on Deep Learning}, proceedings={Advanced Hybrid Information Processing. 6th EAI International Conference, ADHIP 2022, Changsha, China, September 29-30, 2022, Proceedings, Part II}, proceedings_a={ADHIP PART 2}, year={2023}, month={3}, keywords={Deep learning Social network Fake news Communication detection Detection method News communication}, doi={10.1007/978-3-031-28867-8_35} }- Yandan Lu
Hongmei Ye
Year: 2023
Detection Method of Fake News Spread in Social Network Based on Deep Learning
ADHIP PART 2
Springer
DOI: 10.1007/978-3-031-28867-8_35
Abstract
The current detection of fake news spread in social networks does not consider the correlation between news text and images, resulting in inaccurate detection results. A detection method for fake news spread in social networks based on deep learning is devised. The size of the time period is dynamically adjusted according to the number of news in the time period, and features are extracted uniformly for comments/retweets in the same time period. Preprocess social network news data to ensure that the vast majority of text is covered and controlled within the range of machine computing power. Multi-modal features are mined and constructed from images, texts and user-side information. Modal fusion does not use the addition of residuals, but splices the residuals and attention matrices, and then sends them to the fully connected layer to convert the dimension size, and then Update the modal. The fused feature vector is input into the feedforward network for classification, and the prediction result is obtained. The experimental results show that the design method can improve the detection accuracy, and the deep features it contains can more effectively detect fake media content in social networks.
