Search-based testing techniques using genetic algorithm (GA) can automatically generate test data that achieves high coverage on almost any given program under test. GA casts the path coverage test data generation as an optimization problem and applies efficient search-based algorithms to find suitable test cases. GA approaches scale well and can handle any source code and test criteria, but it still has some degrades when program under test has critical path clusters. This paper presents a method for improving GA efficiency by integrating a constraint solver to solve path conditions in which regular GA cannot generate test data for coverage. The proposed approach is also applied to some programs under test. Experimental results demonstrate that improved GA can generate suitable test data has higher path coverage than the regular one.
CITATION STYLE
Dinh, N. T., Vo, H. D., Vu, T. D., & Nguyen, V. H. (2017). Generation of Test Data Using Genetic Algorithm and Constraint Solver. In Studies in Computational Intelligence (Vol. 710, pp. 499–513). Springer Verlag. https://doi.org/10.1007/978-3-319-56660-3_43
Mendeley helps you to discover research relevant for your work.