Misinformation detection in social networks using emotion analysis and user behavior analysis

Social networks have become a prevalent platform for rapid information dissemination among the general populace. However, the proliferation of misinformation has emerged as a critical challenge faced by social network users. Distinguishing between genuine and misleading content poses a significant challenge, leading to inadvertent sharing of misinformation by the majority of users who perceive the information as authentic. Users' emotions, conveyed through their shared posts, play a substantial role in the propagation of misinformation. While several methods leveraging emotion analysis have been developed to detect misinformation in social networks, the influence of user attributes beyond emotion remains relatively understudied. This study aims to explore the role of user attributes in conjunction with emotions to identify misinformation in social networks. Emotions expressed in tweets are extracted and analyzed using RNN, and the likelihood of misinformation in tweets is determined using a Fuzzy Inference System. Six fundamental features extracted from users' social media profiles, coupled with emotions identified from user-posted text, are employed to discern the presence of misinformation in tweets. The proposed mechanism has been tested across five real-world datasets, demonstrating that combining emotion analysis with user behavior serves as a reliable indicator for detecting misinformation.