From ontologies to input models for combinatorial testing

Abstract

Ordinary tools for computing combinatorial test suites rely on simple input models comprising variables together with their domains and constraints limiting possible combinations. Modeling for combinatorial testing requires to represent the input domain of the application in a way such that it fits to the combinatorial testing input model. Depending on the application’s domain this mapping ranges from trivial to more complicated. In this paper, we focus on modeling for combinatorial testing in cases the application’s domain can be represented in form of an ontology, i.e., concepts and their relationships. We formally introduce the notation of ontology we rely on in this paper, and show how such ontologies can be automatically mapped to a combinatorial testing input model. We discuss the algorithm and show its properties.