Power genesis in social networks: An entropy-driven decision support model with conditional data

Abstract Power is a central phenomenon in human societies. Sociologists, politicians, philosophers in all cultures try to understand emergence and change processes of power. What was missing for a long time is measurement of power. First attempts were made in the last decades. This contribution generalizes: It proposes a behaviorist model to measure power in an entropy-driven, conditional-probabilistic framework. For this, relational data must be collected to model the structure of the social fabric. Then, power indices of all actors and groups of actors can be determined and power changes can be simulated to support decision-making on e.g. network design/dismantling or information spreading. The new concept is applied to a group of criminal clans and their mutual relationships, leading to a new network design where a formerly weaker clan becomes the most powerful one after reorganization.