Prioritization of software test cases: an approach using "Clustering" and AHP-Gaussian

The present study aimed to order the software test cases in order to Get an execution priority that will maximize the rate of fault detection. For this job Two approaches were used, one using the "Analytic Gaussian Hierarchy Process" [AHP-Gaussian] and another using the clustering method "K-Means" in conjunction with the AHP-Gaussian. In the formation of the clusters, it was used Textual similarity of the functional objective of each test case to the algorithm "word2vec". And for the criteria of the AHP-Gaussian method, the number of failures was considered, the number of steps and the average execution time of each test case. As a metric for In order to evaluate the performance of each approach, the "Average Percentage of Fault Detection" was used. [APFD]". And the AHP-Gaussian + Clustering method proved to be more effective, presenting a Highest average percentage of the fault detection rate value.