Combining Agent-Based and Social Network Analysis Approaches to Recognition of Role Influence in Social Media

These days, different forms of social media play a significant role in the functioning of individuals and society, and social network analysis methodology ensures a better understanding of the structure and behaviour of societies forming in such environments. Including an agent–based approach to such analyses allows a more complete understanding of the specificity of given users as well as the local interactions between them. In this paper we introduce a multi–agent model of user organisations in social media, and analyse roles in social organisations based on distinguishing features of user behaviour such as activity, cooperativeness and group formation. We also analyse the range of influence of users playing given roles in the society, taking into consideration the consequences of removal of users with specific roles and carry out several experiments with data from the political blogosphere.