Organizational models within multi-agent systems literature are of a static nature. Depending upon circumstances adaptation of the organizational model can be essential to ensure a continuous successful function of the system. This paper presents an approach based on max flow networks to dynamically adapt organizational models to environmental fluctuation. First, a formal mapping between a well-known organizational modeling framework and max flow networks is presented. Having such a mapping maintains the insightful structure of an organizational model whereas specifying efficient adaptation algorithms based on max flow networks can be done as well. Thereafter two adaptation mechanisms based on max flow networks are introduced each being appropriate for different environmental characteristics.
Subjects: 7.1 Multi-Agent Systems; 7. Distributed AI
Submitted: Oct 3, 2006