UGC-driven social influence study in online micro-blogging sites

In Web 2.0 era, the content on a web page is increasingly generated by end users, rather than limited number of administrators. Hence, large number of User Generated Content (UGC) has driven the explosion of content in the web. Thanks to UGC, the pattern of web usage has evolved from download dominated way to a hybrid one with both information download and upload. Large number of UGC has unveiled great capacity of information that is unavailable for researchers before, such as individual preferences, social connections, and etc. In this paper, we propose a novel model which studies the UGC in micro-blogging web sites, the largest and fastest information diffusion media online, and evaluate the social influence for an arbitrary individual. Experimental results show that our model outperforms state-of-the-art techniques in social influence evaluation in both the running time and accuracy.