Experiments in parallel execution of answer set programs

In this paper we study the use of parallelism to speed up execution of Answer Set Programs (ASP). ASP is an emerging programming paradigm which combines features from constraint programming, logic programming, and non-monotonic reasoning, and has found relevant applications in areas such as planning and intelligent agents. We propose different methodologies to parallelize execution of ASP programs, and we describe a prototype which exploits one of such forms of parallelism (vertical parallelism). Performance figures are presented and analyzed along with various considerations regarding scheduling and optimization.