The science of twitter lists: Understanding membership and subscription through network analysis

We report on an exploratory analysis of membership and subscription patterns in publicly shared Twitter lists. Our analysis is conducted in two phases. In the first phase, we examine membership and subscription patterns of lists by defining specific quantitative measures. In the second phase, using network analysis, we propose the use of structural holes to assess the assortativity of a network. We examine the partial structure of the network around a node (depicting a list member/subscriber) to discover users' implicit preferences in subscription and membership. We find that distribution of number of list subscribers follows a power law with most lists having very few or no subscribers. We also find that Twitter users usually segregate groups of people and distribute them across their public lists in such a way that there is a very little overlap among the lists. We find similar results in the case of members of lists to which a user has subscribed. We also show that the structure of the network around a list and its curator can help us understand implicit preferences in subscription and membership. Finally, we find that the network characteristics of a list can help predict churn in subscribers more accurately than churn in members.