The evolution of online forums as communication networks: An agent-based model

This article proposes a theoretical and computational model of how people communicate in online discussion forums. The model's analysis shows that two simple mechanisms can account for the complexity of online communication networks, i.e., a) the tendency to talk with already known users, and b) the tendency to engage in Q&A exchanges. These two mechanisms are able to reproduce lengthy discussions as well as complex interactions between users with different levels of expertise. Most notably, when compared to empirical data taken from three different online forums, the model is able to discern when a strong core of users emerges (or not) that interacts more frequently. Since the evolution of a core of stable interacting users might be the first step in building a community's shared feeling, being able to model the emergence of such core is of interest in a broad range of contexts, from policy-making and e-learning to social marketing.