The smart distribution of social media contents

Abstract

Today's data such as social media contents and archive of digital contents gathered via ubiquitous devices have been hosted on cloud and being shared in a distribution manner. This causes network link congestions, delayed cloud services and increases in public cloud data-out charges. Simulations have demonstrated that deploying our approach, i-Cloud, as the core mechanism of cloud cache could alleviate these problems up to 17.24% byte-hit and cost-saving, 17.96% delay-saving and 29.33% cache hit outperforming the other well-known approaches. A main finding was that there is no significant performance difference between i-Cloud learning single-user-community patterns and i-Cloud learning cross-user-community patterns of comparable sizes.