Mashuk Ahmed creator author. text my 2023 2023 monographic eng xii, 78 pages : illustrations (some color) ; 30 cm. + 1 CD-ROM specialized Mashuk Ahmed Faculty of Computing Thesis (Master of Science) -- Universiti Malaysia Pahang – 2023 In order to save time and resources in software testing, it’s important to avoid exhaustively testing all test cases. For strengthening the software testing system, the t-way testing strategy using metaheuristic algorithm has been useful to process a large number of input combinations for obtaining optimal solutions. However, almost all existing t-way testing strategies mainly consider test case weight while constructing test suites; there has hardly been any consideration of test case priority in the existing works. But due to the continuously increasing complexity of software systems at present, it’s often significant to identify and classify high-priority test cases and low-priority test cases. Focusing only on test case weight is helpful in reducing the test suite size, but it may not be enough for complex software systems. In most of the current t-way testing methods, due to not considering test case priority while testing, many highly prioritized test cases can stay ignored, causing unsatisfactory test outcomes. For test case prioritization while constructing t-way test suites, further improvements on the existing t-way strategies are necessary. Therefore, this research has proposed a t-way testing method implementing an adaptive Dragonfly Algorithm (DA) to construct prioritized test suites. While constructing test suites in this strategy, both test case weight and test case priority have equal importance. The research has also analyzed the potential of the Dragonfly Algorithm for generating prioritized t-way test suites. There has been an implementation of a Bi-objective Dragonfly Algorithm (BDA) to construct prioritized test suites, where the two objectives are test case weight and test case priority. The experimental results demonstrate that the proposed method performs strongly against existing t-way testing methods like PSO, HSS, FPA, All Pairs, VPTag, SCA, and TLBO in term of test suite size, and in addition, the test cases generated by the proposed method have high priority levels. Non-parametric tests like Friedman Tests and Wilcoxon Signed-Rank Tests were conducted for the statistical analysis of the test results. The null hypothesis was rejected 57.14% times, and in the remaining cases, the proposed method showed competitive results compared to other methods. Overall, the designed t-way testing method has been successful in constructing smaller test suites and ensuring high priority levels of constructed test cases. Faculty of Computing Dissertations Universities and colleges Dissertations Theses Dissertations THE0009822 (Local) UMPSA 240108 20251125110825.0 eng