FLACORM: fuzzy logic and ant colony optimization for rumor mitigation through stance prediction in online social networks

Online Social Networks facilitate quicker and larger dissemination of information, but suffer from wide-spread diffusion of rumors. The rumors may influence the users by promoting or discrediting specific products or targets in online social networks. Identification of rumor spreaders while considering different types of rumors and the diversity of the users is an important research challenge. The content of the rumor and the context play a significant role in determining the user's stance on the particular topic. This paper aims at predicting the stance of the users on a rumor and thereby identifying the users who are likely to spread the rumor. The impact of user stance on rumor spread has been investigated in this research work. We propose a Fuzzy logic-based approach to predict the user stance on a rumor based on the content (semantic and contextual features), sentiment, orientation of the rumor and the user's predisposition toward the target of the rumor. Further, a user-stance-based ant colony optimization algorithm (FLACORM) has been proposed to determine the top rumor spreaders. The proposed approach considers the user stance and user predisposition toward the rumor to predict the rumor spreaders accurately. Experiments are carried out to validate and to show the efficacy of the proposed algorithms on real-world datasets.