Cost-effective regression testing through Adaptive Test Prioritization strategies

• We propose two new ATP (Adaptive Test Prioritization) strategies.
• We conduct an empirical study investigating existing and new ATP strategies.
• We provide a statistical analysis examining all ATP strategies proposed.
• Our findings show that FESART is the most consistent cost-effective ATP strategy.

Regression testing is an important part of the software development life cycle. It is also very expensive. Many different techniques have been proposed for reducing the cost of regression testing. However, research has shown that the effectiveness of different techniques varies under different testing environments and software change characteristics. In prior work, we developed strategies to investigate ways of choosing the most cost-effective regression testing technique for a particular regression testing session. In this work, we empirically study the existing strategies presented in prior work as well as develop two additional Adaptive Test Prioritization (ATP) strategies using fuzzy analytical hierarchy process (AHP) and the weighted sum model (WSM). We also provide a comparative study examining each of the ATP strategies presented to date. This research will provide researchers and practitioners with strategies to utilize in regression testing plans as well as provide data to use when deciding which of the strategies would best fit their testing needs. The empirical studies provided in this research show that utilizing these strategies can improve the cost-effectiveness of regression testing.