Inferring social influence and meme interaction with Hawkes processes

Revealing underlying social influence among users in social media is critical to understanding how users interact, on which a lot of security intelligence applications can be built. Existing methods fail to take into account the interaction relationships among memes. In this paper, we propose to simultaneously model social influence and meme interaction in information diffusion with novel multidimensional Hawkes processes. Experimental results on both synthetic and real world social media data show the efficacy of the proposed approach.