Reconstruction of social group networks from friendship networks using a tag-based model

•A tag-based model is proposed to explain the mechanism on the growth of social groups.•In the model, social groups expand on a friendship network based on users’ tags of interest.•Users’ activity in joining group is related to their degree of friendship network.•Various distributions of the simulated group network are in agreement with empirical findings.•Our model throws light on the reconstruction of institute-based relationships.

Social group is a type of mesoscopic structure that connects human individuals in microscopic level and the global structure of society. In this paper, we propose a tag-based model considering that social groups expand along the edge that connects two neighbors with a similar tag of interest. The model runs on a real-world friendship network, and its simulation results show that various properties of simulated group network can well fit the empirical analysis on real-world social groups, indicating that the model catches the major mechanism driving the evolution of social groups and successfully reconstructs the social group network from a friendship network and throws light on digging of relationships between social functional organizations.