On an Argument-centric Persuasion Framework

In this paper, we propose an argument-centric persuasion framework. We first introduce a decision problem, called persuasion satisfiability, which is defined as the problem of determining whether there exists a sequence of arguments that starts from a given initial state, such as beliefs or wishes of the persuadee, and allows for achieving a given purpose of the persuader. This sequence should satisfy different constraints, including particularly upper bound constraints on the weight as well as on the length. We show that this decision problem is NP-complete and propose an encoding in partial weighted MaxSAT framework for solving it. Then, we show that the proposed encoding offers flexibility for dealing with different variants of the persuasion satisfiability problem. Finally, to avoid the explicit use of upper bound constraints on the weight and the length, we consider the notion of Pareto optimality by proposing an approach based on the use of partial weighted MaxSAT, which allows for finding non dominated (optimal) solutions.