Input Modeling Prioritization Using Statistically User Profile for Pairwise Test Case Generation with Constraints Handling

Abstract

Pairwise testing is a widely used technique for software testing with the reduced size of the test suite and able to detect interactions that trigger the system’s faults. In addition, pairwise testing test suites must be able to take constraints between input parameters and parameter values into account. In current practice, identifying and selecting input parameters and parameter values usually depends on tester skills that might not be sufficient. Input parameters and parameter values modeling and tools for easily guiding and prioritizing the selection of optimal input parameters and parameter values for the SUT is also required. In this work, we present an approach for prioritizing input parameters and parameter values modeling using statistical user profile. Our approach is implemented in a tool called UPPTCT which provides the ability to handle constraints on input parameters and parameter values for pairwise testing in order to generate test cases. We conduct experiments to evaluate test case effectiveness and compare our tool with other renowned pairwise test generation and constraints handling tools. The experimental results show that the effectiveness of our approach is significantly more efficient and effective than random testing as a large portion of reported defects with regard to statically user profile were caught by our approach. Furthermore, our tool performs better in some cases and performs comparable results for generating test cases upon input parameters and parameter values for both with constraints handling and without constraints handling.