Information diffusion and link creation are the two major types of dynamic processes in online social networks. Although much effort has been done to study each process separately, their coupling existence and interaction have rarely been observed or investigated.In this paper, we reveal our discovery that the creation of new links is usually coupled with precedent information sharing behavior. That is, before a user truly decides to follow another user, he or she usually chooses to retweet from that user first, acting as a non-follower retweetor. Based on a large microblogging social network, we systematically investigate how different factors impact the non-follower retweetors' transformation to future followers. Specifically, we focus on factors stemming from three aspects: non-follower retweet pattern, user demographics, and social effects. Our analysis presents several interesting and brand new discoveries. We fit all the observed factors into a prediction model and further verify the effectiveness of our investigation. We believe our findings provide new perspectives on the underlying mechanism of dynamics in social networks.
展开▼
