On reduced neighborhood mutation analysis using a single mutagenic operator

Abstract

Mutation analysis evaluates the quality of a test suite based on its ability to reveal simple faults, termed mutations, injected in the program under test. Unfortunately, the need to execute many slightly different versions, or mutants, of a program makes mutation analysis a computationally expensive technique. This paper reports on a promising approach that reduces the cost of mutation analysis by reducing the number of mutants that need to be executed.