An output-oriented approach of test data generation based on genetic algorithm

Abstract

Using genetic algorithm to transform test data generation problem into numerical optimization problem, evolution test is one of the hot topics in test data automatic generation. This paper proposed a software test data generation method based on evolution test, which was output-oriented and so suitable for black-box testing. The method transformed the coverage to software output domains into coverage to branches of pseudo-path by use of gray-box test technology. It defined a match function to describe the difference of the search trace to the aimed path, and then got its fitness function based on the match function. Some experimental results showed that the method implemented the coverage to software output domains, and was more efficient than random testing and manual testing.