Advanced independent cascade model for YouTube content propagation in Facebook

Multimedia streaming service providers such as YouTube often need to design caching strategy based on predicted future media content propagation pattern. It is commonly observed that social network behavior tends to have great correlation with the content propagation pattern. In this paper, we propose an Advanced Independent Cascade Model (AICM) which is an extension of the existing ICM method to predict the future propagation pattern of multimedia (YouTube) content based on scraped social network (Facebook) user profiles in terms of their “talkativeness” and “influential power”. Simulation results suggest that the proposed AICM is reasonable and consistent.