Requirement-based Selection Model for the Expansion of Regression Test Selection

Abstract

Issue of maintaining software is to consider which test cases should be kept for the next modification where the size of test suite gets bigger. This makes performance of software development pull out. The objective of proposing requirement-based test case selection model is to improve ability of regression test selection, in particular, to moderate the size of test suite of the modified program, which gets larger after modification regarding the need of specific requirements, including preparing higher ability of removing faults. It comprises five main algorithms, which are finding reused test case, classifying, revising, deleting, and selecting the appropriate test cases. This paper uses six programs run on different four comparative studies, which are selectall, random, and regression test selection. It gives smaller size than the traditional techniques about 49.78% in average. Besides, it offers percent fixing faults that is higher than select-all, random, and regression test selection algorithm as around 0.06–1.32%.