An application of program slicing in evolutionary testing

Abstract

Evolutionary Testing is a kind of efficient method of automatically test data generation. It uses a kind of metaheuristic search technique, the Genetic Algorithm, to convert the task of test data generation into an optimal problem. One problem for evolutionary testing is the convergence speed will be degression when a problem has many variables and is within a large input domain. In this paper, we suggest that program slicing can be used to narrow the search space, and the test data generation framework that applies the method is given, and then a case study is presented in the end. ©2009 IEEE.