Using Mutation Analysis to Evolve Subdomains for Random Testing

摘要

Random testing is inexpensive, but it can also be inefficient. We apply mutation analysis to evolve efficient sub domains for the input parameters of eight benchmark programs that are frequently used in testing research. The evolved sub domains can be used for program analysis and regression testing. Test suites generated from the optimised sub domains outperform those generated from random sub domains with 10, 100 and 1000 test cases for uniform, Gaussian and exponential sampling. Our sub domains kill a large proportion of mutants for most of the programs we tested with just 10 test cases.