Kaynak kod değişimlerinden ilgili test senaryolarının bulunması

Software testing is commonly used for validating software changes, but it is the most expensive phase of the software development life cycle (SDLC). Test case prioritization (TCP) aims to reduce the cost by scheduling the running order of tests to increase the effectiveness of testing; so that most beneficial test cases are executed first, and faults are detected in the early phases of testing. In the study, we present a novel static TCP technique for manual black-box testing. We use a topic modeling (TM) algorithm to extract the functionalities of each test script. This approach allows differentiating and ranking test cases. Cases those test different parts of the system under test (SUT) get higher ranks. Our approach is compared with the manually prioritized test cases of test engineers in the case study of a commercial online banking project. The comparison shows that the average percentage of fault detection (APFD) rates of our approach is higher than the manual prioritization approach.