Towards mutation analysis of Android apps

摘要: Android applications (apps) have the highest number of releases, purchases, and downloads among mobile apps. However, quality is a known problem, hence there significant research interest in better methods for testing We identify three reasons to extend mutation First, current approaches apps use simple coverage criteria such as statement coverage; extending promises more sophisticated testing. Second, researchers inventing other test need evaluate their selection strategies, which excels at. Finally, some generation apps, specifically combinatorial approaches, generate very large numbers tests. This particularly problematic because running tests slow. For these reasons, this paper proposes an innovative analysis approach specific define operators characteristics extensive XML files specify behavior. implemented prototype tool generating, installing, executing mutants on systems. report preliminary results that show feasible we challenges be addressed effective.