A test case selection framework and technique: weighted average scoring method

Abstract

Test case selection techniques identify and eliminate the modification revealing test cases and try to reduce the test suite size for optimization of regression testing. The objective of this experiment is to assess the effectiveness of weighted average scoring method of test case selection against single objective test case selection techniques. The multi-objective test case selection with the weighted average scoring framework and technique are proposed in this study to select the test cases. This method is trying to solve conflicting test case selection objectives with six selection scenarios. The method used test data of cost, coverage, fault detection ability and code change information, convert them into the weighted average score as scalar function and presented this score to 100-index slabs related to low to high scores, then select the test cases. The results for these selection scenarios are computed and evaluated using size reduction of the test suite, inclusiveness, and precision. The results showed that all scenarios performed acceptable level within conditions applied from 17% minimum to 86% maximum in size reduction metrics. The inclusiveness showed 17% to 88% and 33% to 85% for precision metric.