Elsevier

Artificial Intelligence

Volume 275, October 2019, Pages 379-410
Artificial Intelligence

A general notion of equivalence for abstract argumentation

https://doi.org/10.1016/j.artint.2019.06.006Get rights and content
Under an Elsevier user license
open archive

Abstract

We introduce a parametrized equivalence notion for abstract argumentation that subsumes standard and strong equivalence as corner cases. Under this notion, two argumentation frameworks are equivalent if they deliver the same extensions under any addition of arguments and attacks that do not affect a given set of core arguments. We also provide exact characterizations and complexity results. The proposed notion of equivalence is motivated by its capability to capture the concept of local simplifications. In fact, our equivalence notion allows to decide whether a sub-framework can be replaced by another one without changing the extensions in the framework which undergoes this change. Moreover, as our characterizations demonstrate deciding this form of equivalence does not require an analysis of the entire framework. This makes it an appealing formal underpinning for establishing general replacement patterns in argumentation frameworks.

Keywords

Abstract argumentation
Equivalence
Local simplification
Computational complexity

Cited by (0)