Analyzing cross-platform information propagation.

Summary Using data gathered from multiple sources across the World Wide Web (social networks, news websites, etc.), the objective of this work is to describe the features behind the information propagation on the web. We first group the collected data by their similarity that may have been induced by a common factor, e.g. similar topic. Second, we go through information cascades analysis in terms of their temporal and structural characteristics that allows us to find their main propagation patterns. Third, we investigate the problem of clustering the cascades into groups that behave similarly with respect to their diffusion on the web. The presented approach is unsupervised and uses only behavioral data.