Enhanced classification tree method for modeling pairwise testing

Abstract

Software testing is one of the most important activities to produce a high-quality system, which can increase the trust level of users. There are many types of software testing. One of those testing is called exhaustive testing. Exhaustive testing is used to produce a test suite that will be used in other testing types such as unit testing, system testing, integration testing and also acceptance testing. However, exhaustive testing is infeasible and will be time consuming. Therefore, the combinatorial testing is proposed to solve the exhaustive testing problem. There are many techniques of combinatorial testing. The popular one is called pairwise testing. It also is known as Allpairs or 2-way testing. It involves the interaction of 2 parameters. In order to perform the pairwise testing, there are procedures that need to be fulfilled. The first procedure is modeling of System Under Test (SUT). There are many models that can be used to design the test suite for pairwise testing. In this paper, the comparison for modeling of SUT in pairwise testing is performed, and the enhancement of Classification Tree Method is proposed. An example based on steps of proposed model method is also provided.