Deliberation About Preferences and Group Decisions

Our research consists of an approach to Group Decision Making in multiagent systems that is inspired by human interaction. The main goal of the thesis is to evaluate how engaging in deliberation can help a group of agents to improve their decisions. In order to achieve that, we proposed a model of preferences that allows arguing about them and a protocol of deliberation which is inherently collaborative, but where preferences play a role to render agents individual and collective opinions. Our next step is to find metrics to quantify the satisfaction of an agent towards a decision, so we can compare the decisions in our method against those provided by social choice methods.