A Comparison of Heuristics for Answer Set Programming

Answer Set Programming (ASP) is a novel programming paradig m, which allows to solve problems in a simple and highly declarative way . The language of ASP (function-free disjunctive logic programming) is very expressive, and allows to represent even problems of high complexity (every proble m in the complexity class P2 = NPNP). As for SAT solvers, the heuristic for the selection of the bra nching literal (i.e., the criterion determining the literal to be assumed true at a given stage of the computation) dramatically affects the performance of an ASP sy stem. While heuristics for SAT have received a fair deal of research in AI, only littl e work in heuristics for ASP has been done so far. In this paper, we extend to the ASP framework a number of heuri stics which have been successfully employed in existing systems, and we compare them experimentally. To this end, we implement such heuristics in t he ASP systemDLV, and we evaluate their respective efficiency on a number of ben chmark problems taken from various domains. The experiments show interesti ng results, and evidence a couple of promising heuristic criteria for ASP, whic h sensibly outperform the heuristic ofDLV.