SCORE: A Social-Aware Content Distribution Method for Social Networks

The generation and share of user generated content via online social networks (OSNs) aggressively promotes the substantial growth of Internet traffic. When confronting such massive network traffic, how to improve quality of experience for users remains an important yet challenging problem. Though there exist works on intelligent distribution and caching of user generated content, none of them comprehensively exploits the valuable social information of OSNs. In this paper, we propose SCORE, a social-aware content distribution method by leveraging the social relationship and user geographic information widely available in most OSNs. In particular, we design distribution content selection and target server selection schemes to effectively push content to geo-located servers. We evaluate SCORE using real-world social network traces. Experimental results demonstrate that SCORE can significantly reduce the future user access latency and alleviate network traffic pressure.