Abstract:
Fault Seeding is a testing technique where faults are artificially injected into an application to assess the effectiveness, i.e. if a given test suite is capable of uncovering the injected faults, of a test suite. This is helpful in establishing confidence in the test suite and is an alternative to structural testing methods. One of the issues with fault seeding is the identification of potential areas in the application, where the faults are to be seeded. We argue that if the intended usage of the application under test could be inferred from the potential users’ interactions with the application, such information could be incorporated into the fault-seeding process. This could lead to more effective faultseeding in a test application. In this work, we study fault seeding mechanisms based on user interactions with the application; and thus give a guided fault seeding mechanism for the purpose. We show the usefulness of the guided fault seeding with the help of a case study using the blackboard application.
