A general framework for preferences in answer set programming

Abstract

We introduce a general, flexible, and extensible framework for quantitative and qualitative preferences among the stable models of logic programs. Since it is straightforward to capture propositional theories and constraint satisfaction problems with logic programs, our approach is also relevant to optimization in satisfiability testing and constraint processing. We show how complex preference relations can be specified through user-defined preference types and their arguments. We describe how preference specifications are handled internally by so-called preference programs, which are used for dominance testing. We also provide algorithms for computing one, or all, preferred stable models of a logic program, and study the complexity of these problems. We implemented our approach in the asprin system by means of multi-shot answer set solving technology. We demonstrate the generality and flexibility of our methodology by showing how easily existing preference languages can be implemented in asprin. Finally, we empirically evaluate our contributions and contrast them with dedicated implementations.