Identifying the End-to-End Backbone of Information Diffusion

In recent years, online social networks (OSNs) have undoubtedly become the most popular information diffusion channel. However, existing work has rarely focused on the end-to-end diffusion path, which means information diffusion between user pairs is still a problem unresolved. In this paper, an end-to-end independent cascade model (EICM) is proposed. Then, we present an efficient approach called end-to-end diffusion flow pruning (EDFP) that simplify large-scale social networks until main diffusion paths of end-to-end networks are identified.