A preference-based approach to backbone computation with application to argumentation.

The backbone of a constraint satisfaction problem consists of those variables that take the same value in all solutions. Algorithms for determining the backbone of propositional formulas, i.e., Boolean satisfiability (SAT) instances, find various real-world applications. From the knowledge representation and reasoning (KRR) perspective, one interesting connection is that of backbones and the so-called ideal semantics in abstract argumentation. In this paper, we propose a new backbone algorithm which makes use of a "SAT with preferences" solver, i.e., a SAT solver which is guaranteed to output a most preferred satisfying assignment w.r.t. a given preference over literals of the SAT instance at hand. We also show empirically that the proposed approach is specifically effective in computing the ideal semantics of argumentation frameworks, noticeably outperforming an other state-of-the-art backbone solver as well as the winning approach of the recent ICCMA 2017 argumentation solver competition in the ideal semantics track.