RPI researchers Boleslaw Szymanski (PI) and Gyorgy Korniss (co-PI) joined the team of UIUC researchers: Tarek Abdelzaher (PI), Jiawei Han (co-PI) and David Nicol (co-PI) in preparing a proposal titled SocialCube: A Multiscale Modeling and Simulation Framework for Social Streams. In June, this proposal has been selected for funding by the new DARPA program called Social Sim. The goal of this project is enable "what-if" simulations of social media that focus on predicting information propagation in online social environments and on co-evolution of such propagation with changes in individuals' beliefs. The proposed work will advance fundamental knowledge of social media simulation by offering (i) composable models of information propagation on social media that account for interactions among multiple factors, such as group bias, source in influence, and content types, (ii) maximum-likelihood estimation algorithms that extract needed simulation model parameters from social media observations of a target population, (iii) models that predict co-evolution of social beliefs with information dissemination, (iv) multi-granularity solutions that tailor the complexity of simulation models to the specific scenario being simulated, such that desired simulation fidelity is achieved, while attaining scalability, and (v) analytic foundations to quantify accuracy of results. These advances will be applied to multiple different environments, specifically Twitter and Instagram, as well as other media.  

Social Cognitive Networks Academic Research Center