Abstract
Undoubtedly, blogs are a significant representative in the Web 2.0 age. Every user on the Internet is able to own his or her blog space as a platform to publish personal thoughts, ideas, comments or works. Information can thus be propagated and shared in a much easier way and this has led to the development and evolution of user communities in the blogosphere. In this work, the phenomenon of information diffusion is studied through the observation of evolving social relationships among bloggers. Specifically, it is noted that a social network can be formed with bloggers and corresponding subscription relationships. Moreover, the subscription of a blogger to another blog site may reflect that the distance among relevant bloggers is shortened due to information diffusion. Consequently, we devise in this work a new data mining capability, i.e., the shortcut patterns, and a proper approach to identify such patterns. Empirical studies show that the proposed scheme is feasible in various applications, including viral marketing and personalized recommendation.