THE USE OF EMERGENT BEHAVIOUR IN A MULTI-AGENT SYSTEM TO DRIVE SELF-ADAPTATION AT THE INTERFACE

摘要

Multiple simple agents have been used to drive adaptive behavior in a system that presents data in various graphical and tabular forms. Agents observe the users' actions and review the data that is input into the system. Based on their observations, the community of agents make decisions about which display formats to recommend when new data is loaded. Rather than carrying out high-level decision-making, the agents work as an emergent system where the result of their interactions provides the set of recommended displays. This approach has been deployed in the real world domain of medicine.