LAUSR

With the rapid development and growth of online social networks (OSNs), researchers have been pushed forward to improve the knowledge of these complex networks by analyzing several aspects, such as the types of social media, the structural properties of the network, or the interaction patterns among users. In particular, a relevant effort has been devoted to the study and identification of cohesive groups of users in OSNs (also referred as communities) because they are the basic building block of each OSN. While several research works on groups in OSNs have mainly focused on identifying the types of groups and the contents created by their members, the analysis of internal organizations of such groups remains unexplored due to the lack of real data sets containing information about such groups, about their members, and the interactions among them. In this article, we compensate for this shortcoming by studying the main properties of groups defined in OSNs, taking as reference use cases 40 real Facebook groups of different categories that account for a total of about 500.000 users. In particular, we exploit interaction patterns among users and social network analysis to uncover interesting aspects related to the internal organization of groups. Experimental results reveal that the majority of the collected groups exhibit an internal structure where members can be clustered in four subgroups according to the level of tie strength of the relations they have. Furthermore, clusters identified on Facebook groups can provide relevant information about the importance of users within such groups.