Using memetic algorithms for test case prioritization in model based software testing

Abstract

Building high quality software is one of the main goals in software industry. Software testing is a critical step in confirming the quality of software. Testing is an expensive activity because it consumes about 30% to 50% of all software developing cost. Today much research has been done in generating and prioritizing tests. First, tester should find the most important and critical path in software. They can reduce cost by finding errors and preventing to propagate it in design step. In this paper, a model based testing method is introduced. This method can prioritize tests using activity diagram, control flow graph, genetic and memetic algorithm. Different version of memetic algorithm has been made by stochastic local search, randomize iterative improvement, hill climbing and simulated annealing algorithms. The results show that the using local search methods with genetic algorithm (GA) provide efficiency and produce competitive results in comparison with GA.