Analysis of Testing Approaches for Web Applications Based on Each Choice and Pairwise Methods.

摘要

Time and quality are important factors when determining the proper approach for software testing. A software program can often be used in various environments (different platforms, operating systems, browsers, networks, etc.) and require thorough testing to provide high quality and reliability in different configurations. Combinatorial testing is an effective approach to testing hardware and software configurations. However, testing resources are often restricted in real practice. Because business goals require different testing methods, there is no best one-size-fits-all testing approach. For this reason, we experimentally investigated and analyzed several combinatorial approaches based on Each Choice and pairwise methods (with and without the consideration of operational profiles) through the testing of an Adviser Scheduling application located in a university web portal. Test sets with various configurations were generated according to six different combinatorial strategies. The Advanced Combinatorial Testing System (ACTS) tool, which was provided by the National Institute of Standards and Technology (NIST), was used to generate pairwise test sets automatically. The case study web application was retested for each of the proposed testing approaches, and the results were compared after taking into account the number of test cases and the corresponding detected faults. Based on this analysis, we answer three research questions and provide recommendations for the selection of testing approaches to align with different business goals. The recommendation chosen for web applications allowed for improved quality and reduced time for software testing.