We are currently analyzing the typical structure of failure-causing inputs within the input domain. A common assumption is that failure-causing inputs are clustered. This has, however, not yet been shown. If this is the case, approaches that make use of this property can be much more effective. One such approach is Adaptive Random Testing (ART) that we are also investigating.



  1. Christoph Schneckenburger, Johannes Mayer:
    Towards the Determination of Typical Failure Patterns. In: Proceedings of the Fourth International Workshop on Software Quality Assurance (SOQUA 2007), ACM Press, New York, NY, USA, 2007, pp. 90-93.