Test case reduction by using fuzzy clustering with new measure similarity

Software testing is a process which is used for determining the accuracy, completeness and quality of advanced computer software. Software tests via a set of inputs named test case. Test case is set of operations which are implemented in order to consider one of the special features or application software. In the designing test case, redundant test cases are created without any utilization that it is very costly and time-consuming for test unit. Our purpose in this article is to reduce time spend on testing by reducing the number of test cases with fuzzy clustering algorithm which has new measure distance similarity. Because may have more efficient and accurate results. Fuzzy clustering helps test cases that have similar feature to classify, so in this case we can avoid spending time on unnecessary test cases. In addition for more efficient fcm clustering we used new measure similarity which combines from two measures that had less objective function compare to Euclidean and for more evaluate we use it in k means and fcm clustering and in the end we compare it with three measures (Euclidean, Manhattan, Minkowski) which has less objective function compare to Euclidean.

Keywords:

reducing the test case , software testing , fuzzy clustering , Cyclomatic complexity , Standard deviation , measure similarity