A Fuzzy Logic-Based Opinion Update Model for Online Social Networks with Influential Social Bots

Social media is a platform where people express their opinions through user-generated text. Investigating opinion changes of people based on the influence of legitimate users and bots over time is crucial. The influence of bots on opinion evolution can be different from the influence of individuals in a network since the opinion of bots does not evolve over time, irrespective of the opinion of the rest of the network. This work aims at investigating the influence of social bots on users' opinions by considering the user's bias and interactions with other users and bots over time. A fuzzy logic-based approach is designed for opinion updates, allowing for uncertainty in the opinion. By considering the tweet's sentiment, we develop a neighbor-biased-fuzzy opinion influence model and analyze the effect of bots in social media. Experiments are conducted on the Social-Honeypot Twitter dataset to evaluate the performance of the proposed approach. The change in user opinion in the presence of different bots is analyzed. Further, the shift in public opinion over time due to the influence of bots is predicted.