Self-organised system for sustainability and beyond
The aim of the NORMS4SRA (Norm-governed self-organised systems for sustainable resource allocation) project was to develop a probabilistic rule-based argumentation framework for investigating norm-governed self-organised systems using two lines of research. The first line of research created theoretical models for theoretical agents to forecast how people would behave under certain circumstances and in a particular technical infrastructure. The second approach designed operational models for building norm-governed self-organising systems with particular properties. Researchers developed a layered framework whereby each one of six layers addressed different requirements. The first two layers used an intuitive rule-based argumentative logic to facilitate communication and update models of the systems under study. The third integrated probability theory into the argumentative framework in order to capture uncertainty. The fourth layer accounted for the learning aspects of the framework, which allowed learning agents to be investigated and models to learn from facts. The fifth focused on norms that controlled or emerged in societies of learning agents. Finally, layer number six dealt with societies where agents governed themselves, by using their own experiences to guide decision-making. Researchers built a proof-of-concept multi-agent layered framework to investigate populations of norm-based agents. This enabled rapid prototyping while accurately representing and communicating models of agents, norms and reason to the framework. The framework was tested by applying it to resource allocation and the field of electrical grids. NORMS4SRA offered an alternative to the current equation-based and game theory models used to study law and economics. It will allow greater insights into complex normative systems, thereby helping policymakers to draw up the appropriate rules to achieve their goals. The project also has wider implications for society. These range from next-generation legal expert systems to reasoning engines for the Internet of Things as a result of on-the-fly learning patterns of events and giving explanations for events.
Keywords
Self-organised system, norms, sustainable resource allocation, argumentation framework, learning agents
