Computing Preferred and Weakly Preferred Answer Sets by Meta Interpretation in Answer Set Programming

Preferred and Weakly Preferred Answer Sets are extensions to Answer Set Programming (ASP) which allow the user to specify priorities for rules. In this paper we present a first implementation of these formalisms by means of “metainterpreters” on top of DLV, an efficient engine for Disjunctive ASP. This approach shows the suitability of ASP in general and of DLV in particular for fast prototyping and experimenting with new languages and knowledge-representation formalisms. In addition to two “straightforward” metainterpreters, we also present a graph-based meta-interpreter that often allows for more efficient computations.