Evaluating several path-based partial dynamic analysis methods for selecting black-box generated test cases

Many systematic methods help to produce a comprehensive repository of test cases that contain no redundancy from the specification point of view. When the repository cannot be tested in its entirety owing to resource constraints, there is a need to select a smaller subset of test cases that are effective with respect to a given implementation. One way of doing so is to select only one from each pair of test cases that are judged to be processed similarly by the given implementation, as determined by partial dynamic analysis. This work examines several path-based partial dynamic analysis methods for the selection of specification-based test cases, and evaluates their effectiveness by experiments.