Preference-based argumentation built from prioritized logic programming
Toshiko Wakaki · Journal of Logic and Computation · 2013
In this article, we introduce a new approach of preference-based argumentation frameworks. We first introduce an abstract preference-based argumentation framework (an abstract PAF) which handles static preferences. Naturally extending Dung's acceptability semantics, the semantics of an abstract PAF is given by P-extensions which ensure conflict-freeness w.r.t. the attack relation. Second, as the instantiation of the proposed abstract PAF, we present the non-abstract PAF along with the constrained argumentation framework (CAF) built from a prioritized logic program (PLP). As a result, the proposed framework is able to handle static preferences as well as integrity constraints in argumentation. Third, by providing a hierarchical PLP to deal with dynamic preferences in the context of logic programming, we extend the framework to a dynamic PAF built from a hierarchical PLP to deal with dynamic preferences in the context of argumentation. Fourth, we not only show the correspondence between stable extensions of the non-abstract argumentation framework and paraconsistent stable models of an extended logic program but also show the correspondence between the semantics of such PAFs and the semantics of the underling prioritized logic programs based on the order-preserving mapping functions. Finally, the proposed framework is also applied to interactions between agents in multi-agent systems.