Positive and Negative Testing with Mutation-Driven Model Checking

Abstract

Mutation-driven test case generation with model checking has been pro- posed to reduce the costs of specification-based mutation analysis. Most of the exist- ing work focuses on verifying the expected behavior in the original model, i.e. positive testing. In this paper negative testing is introduced to check the unexpected behavior. Mutants are divided into three types: increscent, decrescent, and cross mutants. Both, positive and negative testing is proposed to guarantee the detection of these mutants. A non-trivial example illustrates and validates our approach.