Rumor Propagation Detection in Complex Networks Based on ILSR Model and Nodes Degree

The pervasive nature of social networking platforms has led to the production of large amounts of data. The lack of restrictions on information sharing on these networks allows information to be spread regardless of their validity. Such misinformation usually leads to the production and dissemination of rumors. Therefore, automatic detection of rumors on social networks is one of the attractive research areas for the analysis of these networks. This paper introduces a way to deal with the spread of rumors on social media based on the autoencoder neural network model and the ILSR epidemic dissemination model. Here, the autoencoder neural network model is applied with several innovative thresholds for initial rumor detection and then rumor control is performed by an extended version of the ILSR epidemic outbreak model. The proposed model called ILSHR considers social network users in five groups: Ignorant, Lurker, Spreader, Hibernator, and Removal. According to the transmission modes in the ILSHR rumor model, in addition to the characteristics of the lurker group, this model also considers the characteristics of the hibernator group of the SIHR model. Mechanisms of forgetfulness and reminders of hibernation can delay the rumor terminal time and ultimately reduce the impact of the rumor on the social network. The proposed method for modeling the spread of rumors has been performed on the Sina Weibo social network dataset. The results show better performance of the proposed method with 95.7% detection accuracy compared to DGRU and DLSTM models.