A distributed normative infrastructure for situated multi-agent organisations

Abstract

In most of the existing approaches to the design of multiagent systems, there is no clear way in which to relate organisational and normative structures to the model of the environment where they are to be situated and operate. Our work addresses this problem by putting together, in a practical approach to developing multi-agent systems (and social simulations in particular), a high-level environment modelling language that incorporates aspects of agents, organisations, and normative structures. The paper explains in some detail how the ideas of normative objects and normative places, put together as a distributed normative infrastructure, allow the definition of certain kinds of situated multi-agent organisations, in particular organisations for multi-agent systems that operate within concrete environments. Normative objects are environment objects used to explicitly convey normative content that regulate the behaviour of agents within the place where such objects can be perceived by agents. The paper briefly introduces such concepts, showing how they were integrated into the MASSOC multi-agent systems platform for social simulation, and hints on new problems of (situated) organisational and normative structures that were brought forward by the work presented here.