Analyzing online news dissemination via structure learning: An experimental view

Online information dissemination has attracted unprecedented attention with the proliferation of Internet. This paper investigates how news is disseminated through key online media. Key media are defined to include two categories: a) leader media, whose reports will be reproduced by numerous other media; b) source media, serving as the information counselor for leader ones. Through analyzing the appearance of the same report on various online media, we are able to locate key media in news dissemination and predict the path of dissemination. We provide the initial experimental results on real-life datasets, and the results presented in the form of Bayesian network indicate that the unique influence of online media in three different categories during the process of report.