FINDINGS AND IMPLICATIONS OF TEST CASE PRIORITIZATION TECHNIQUES FOR REGRESSION TESTING

Abstract

Software makes headway with time which results in an expanded number of test cases. Software testing and maintenance consumes 50% cost of the overall software development process. This is imperceptible to re-execute the complete test suite as and necessary to evolve the software keeping in mind the end goal to minimize the cost and efforts. Regression (moving in reverse) is an unavoidable truth in programming frameworks. Despite the fact that something worked previously, there is no certification that it will work after the most recent "minor" change. Regression testing is carried out during maintenance phase to confirm whether faults are eliminated or not after the software experiences alterations or modifications and ingest 80% cost. It is the technique to select, minimize and prioritize test cases with the specific end goal to permit testers to detect faults as early as possible during the maintenance phase. So, optimizing regression testing is the prime objective of any system testers with a specific end goal to minimize the overall cost. Test case prioritization is a technique to schedule test cases so that test case with higher priority executes first thus, enhancing the performance objective of early fault detection. In this work, we present findings and implications of different code based and customer based test case prioritization techniques and also suggest soft clustering technique for effective cluster formation based on dependency between test case functions and faults experienced.