Paths-oriented Test Data Generation using Genetic Algorithm

Software testing is one of the most important issues in quality assurance of software products. In traditional testing methods, due to the complexity and high cost, full testing is impossible. The search-based approaches are emerging as a solution for the automatic generation of test data in software testing. In this paper, we employ a genetic algorithm based on the execution path to generate the test data. In the proposed method, instead of checking the desirability of a generation for all execution paths, we assigned an appropriate fitness function per each execution path. Thus, the fitness function for each path, identifies an appropriate generation with great precision and less time. Evaluation results show that the proposed method is able to increase the accuracy of the algorithms as well as reduce the generation steps and the process time.