Skip to Main Content
Dynamic Random Testing (DRT) is a testing strategy to improve traditional random testing and random-partition testing strategies. The essential idea of DRT is that the testing profile, based on which test classes are chosen to select next test case, should be dynamically adjusted according to the testing history. DRT can have various forms depending on parameters settings, application scenarios and test suite partitioning schemes. This paper discusses the theoretical boundary of the parameters in DRT and their impact on the actual effectiveness in practice. A sufficient condition for the parameters estimation in DRT is proposed and further examined through experiments on three subject programs. Data indicates that the proposed sufficient condition provides a reasonable guideline for choosing parameters when applying the DRT strategy.
Date of Conference: 18-22 July 2011