Analyzing Information Proliferation and Opinion Evolution in Social Networks using Nature-Inspired Algorithms

Abstract

Information proliferation and opinion analysis in social networks play a vital role to attract the public and government views. Most of the previous works in the literature have discussed information proliferation and opinion analysis separately. In this paper, a new methodology, for analysing both diffusion evolution and opinion dynamics is proposed. The proposed model is built using a forest fire algorithm, cuckoo search and fuzzy c-means clustering. The nature-inspired forest fire algorithm is used to determine the diffuser and non-diffuser of the information in the social networks and the time-stamp based fuzzy c-means clustering with the cuckoo search optimization algorithm is proposed to classify the content of the social network namely twitter into various opinions categories and to determine the opinion dynamics for time. The comparison of results on the different tweets data sets shows that the proposed model could improve the opinion dynamics prediction performance with the measurements namely precision, recall and accuracy than the other existing models.