An enhanced weighted method with unique priority value for test case prioritization in regression testing

Regression testing is an important and expensive strategy in software testing. To reduce its cost, many methods were proposed including test case prioritization methods. The aim of prioritization methods is to define an ideal order of test cases that allows higher coverage and early fault detection with minimal amount of executed test cases. However, the problem with most of the existing test case prioritization methods is the random sorting of test cases when two or more test cases record equal priority values. In this research, an enhanced weighted method using unique priority value UniVal to prioritize test cases is presented. The proposed method combines five code coverage criteria with the order of test cases from the previous execution session to generate unique priority values. In addition, a controlled experiment was executed and results statistically analyzed to evaluate the effectiveness of the proposed method. The results show an improved performance in terms of prioritizing test cases and achieving higher APFD values. In future, a tool to automate the operation of UniVal would be developed and more experiments would be performed