Automatic generation of atomic multiplicity-preserving search operators for search-based model engineering

Abstract Recently, there has been increased interest in combining model-driven engineering and search-based software engineering. Such approaches use meta-heuristic search guided by operators (model mutators sometimes breeders) implemented as model transformations. The design of these can substantially impact the effectiveness efficiency search. Currently, designing is left to person specifying optimisation problem. However, developing consistent efficient search-operator rules requires not only domain expertise but also in-depth knowledge about optimisation, which makes model-based challenging expensive. In this paper, we propose a generalised approach automatically generate atomic multiplicity-preserving for given This reduces effort required specify an problem shields users from complexity implementing mutation operators. We evaluate our with set case studies show that generated are comparable to, some cases better than, manually created at guiding evolutionary towards near-optimal solutions.