Generating Efficient Mutation Operators for Search-Based Model-Driven Engineering

• Published in: Theory and Practice of Model Transformation Vol. 10374; pp. 121 - 137
• Main Author: Strüber, Daniel
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2017; Springer International Publishing
• Series: Lecture Notes in Computer Science
• ISBN: 9783319614724; 331961472X
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-61473-1_9

Software engineers are frequently faced with tasks that can be expressed as optimization problems. To support them with automation, search-based model-driven engineering combines the abstraction power of models with the versatility of meta-heuristic search algorithms. While current approaches in this area use genetic algorithms with fixed mutation operators to explore the solution space, the efficiency of these operators may heavily depend on the problem at hand. In this work, we propose FitnessStudio, a technique for generating efficient problem-tailored mutation operators automatically based on a two-tier framework. The lower tier is a regular meta-heuristic search whose mutation operator is “trained” by an upper-tier search using a higher-order model transformation. We implemented this framework using the Henshin transformation language and evaluated it in a benchmark case, where the generated mutation operators enabled an improvement to the state of the art in terms of result quality, without sacrificing performance.