Argumentation strategies for task delegation

Abstract

What argument(s) do I put forward in order to persuade another agent to do something for me? This is an important question for an autonomous agent collaborating with others to solve a problem. How effective were similar arguments in convincing similar agents in similar circumstances? What are the risks associated with putting certain arguments forward? Can agents exploit evidence derived from past dialogues to improve the outcome of delegation decisions? In this paper, we present an agent decision-making mechanism where models of other agents are refined through evidence derived from dialogues, and where these models are used to guide future argumentation strategy. We combine argumentation, machine learning and decision theory in a novel way that enables agents to reason about constraints (e.g., policies) that others are operating within, and make informed decisions about whom to delegate a task to. We demonstrate the utility of this novel approach through empirical evaluation in a plan resourcing domain. Our evaluation shows that a combination of decision-theoretic and machine learning techniques can significantly help to improve dialogical outcomes. © 2012 Springer-Verlag Berlin Heidelberg.