Modeling memetics using edge diversity

The diffusion of an idea significantly differs from the diffusion of a disease because of the interplay of the complex sociological and behavioral factors in the former. Hence, the conventional epidemiological models fail to capture the heterogeneity of social networks and the complexity of information diffusion. Standard information diffusion models depend heavily on the micro-level parameters of the network like edge weights and implicit vulnerabilities of nodes towards information. Such parameters are rarely available because of the absence of large amounts of information diffusion data. Hence, modeling information diffusion remains a challenging research problem. In this paper, we utilize the peculiar structure of the real-world social networks to derive useful insights into the micro-level parameters. We propose an artificial framework mimicking the real-world information diffusion. The framework includes (1) a synthetic network which structurally resembles a real-world social network and (2) a meme spreading model based on the penta-level classification of edges in the network. The experimental results prove that the synthetic network combined with the proposed spreading model is able to simulate a real-world meme diffusion. The framework is validated with the help of the diffusion data of the Higgs boson meme on Twitter and the datasets of several popular real-world social networks.